EPA-450/3-77-046
August 1976
            SCREENING STUDY
         TO DETERMINE NEED
FOR SOX AND HYDROCARBON
                  NSPS FOR FCC
               REGENERATORS
  U.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Air and Waste Management
    Office of Air Quality Planning and Standards
   Research Triangle Park, North Carolina 27711

-------
                                   EPA-450/3-77-046
SCREENING STUDY TO DETERMINE
NEED FOR SOX AND HYDROCARBON
   NSPS FOR FCC REGENERATORS
                         by

                    Arthur D. Little, Inc.
                      Acorn Park
                 Cambridge, Massachusetts 02140
                   Contract No. 68-02-1332
                      Task No. 22
               EPA Project Officer: James F. Durham
                      Prepared for

            U.S. ENVIRONMENTAL PROTECTION AGENCY
                Office of Air and Waste Management
             Office of Air Quality Planning and Standards
             Research Triangle Park. North Carolina 27711

                      August 1976

-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers.  Copies are
available free of charge to Federal employees, current contractors and
grantees, and nonprofit organizations - in limited quantities - from the
Library Services Office (MD-35) , Research Triangle Park, North Carolina
27711;  or,  for a fee, from the National Technical Information Service,
5285 Port Royal Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency by
Arthur D. Little,  Inc. ,Acorn Park, Cambridge, Mass. , in fulfillment
of Contract No. 68-02-1332.  The contents of this report are reproduced
herein as received from Arthur D . Little, Inc. The opinions , findings,
and conclusions expressed are those of the author and not necessarily
those of the Environmental Protection Agency. Mention of company or
product names is  not to be considered as an endorsement by the Environment-
al Protection Agency.
                   Publication No.  EPA-450/3-77-046
                           11

-------
                             TABLE OF CONTENTS


                                                                      Page

List of Tables                                                           v

List of Figures                                                        vii

EXECUTIVE SUMMARY                                                     viii

I.    THE PETROLEUM REFINING INDUSTRY                                    1

      A.  EXISTING FLUID CATALYTIC CRACKING UNITS                        1

      B.  GENERAL OPERATING FACTORS                                      1

      C.  GROWTH PROJECTIONS                                            12

      D.  SULFUR CONTENT OF FCC FEED                                    16

II.   SOURCES AND TYPES OF EMISSIONS                                    19

      A.  PROCESS DESCRIPTION                                           19

      B.  REGENERATION                                                  19

      C.  CONDITIONS                                                    22

      D.  EMISSIONS FROM A TYPICAL FCC UNIT                             22

          1.  Gas Volume                                                26
          2.  Particulates                                              26
          3.  SOX                                                       26
          4.  NOX                                                       27
          5.  Carbon Monoxide                                           27
          6.  Hydrocarbons                                              28
          7.  Oxygen                                                    28

      E.  FACTORS AFFECTING SOX EMISSIONS                               28

III.  BEST APPLICABLE SYSTEMS OF EMISSION REDUCTION                     32

      A.  CONTROL STATUS                                                32

          1.  Particulates                                              32
          2.  Carbon Monoxide                                           32
          3.  Oxides of Nitrogen                                        32
          4.  Hydrocarbons                                              33
          5.  Oxides of Sulfur                                          33
          6.  Other Pollutants                                          34
                                     iii

-------
                          TABLE OF CONTENTS cont.

                                                                      Page

      B.  SOX EMISSION CONTROL                                          34

      C.  FLUE GAS DESULFURIZATION PROCESSES                            35

          1.  Elemental Sulfur Processes                                35
          2.  Direct Acid Processes                                     36
          3.  Waste Salt Processes                                      36
          4.  Concentrated S02 Processes                                44

      D.  DESULFURIZATION OF FCC FEED                                   52

      E.  ACHIEVABLE EMISSION LEVELS WITH BEST AVAILABLE CONTROL
          TECHNOLOGY                                                    56

IV.   STATE AND LOCAL AIR POLLUTION CONTROL REGULATIONS                 58

      A.  SELECTION OF STATES FOR REVIEW OF REGULATIONS                 58

      B.  REGULATIONS                                                   58

V.    ESTIMATED EMISSION REDUCTIONS                                     65

      A.  MODEL IV PARAMETERS                                           65

          1.  Parameter Definitions                                     65
          2.  Baseline Year                                             66

      B.  APPLICATION OF THE MODEL                                      68

          1.  Application of TRC Model IV to SOX Emissions              68
          2.  Application of TRC Model IV to Hydrocarbon Emission       83

      C.  IMPACT OF SULFUR AND HYDROCARBON NSPS FOR FCC REGENERATOR     85

VI.   MODIFICATION & RECONSTRUCTION                                     87

REFERENCES                                                              91

APPENDIX A    SUMMARY OF EMISSION DATA                                  93
APPENDIX B    LIST OF CONTACTS                                         112
APPENDIX C    REPORTS AND CORRESPONDENCE                               116
APPENDIX D    BIBLIOGRAPHY                                             145
APPENDIX E    TRC MODEL IV                                             148

APPENDIX F    RELATIONSHIP OF PAD DISTRICTS TO AIR QUALITY
              CONTROL REGIONS                                          170
                                     iv

-------
                              LIST OF TABLES


Table
 No.                                                                 Page

1-1        Location of U.S. Fluid Catalytic Cracking Capacity          2

1-2        U.S. Fluid Catalytic Cracking Capacity by PAD               8

1-3        Capacity of Refineries with FCC Regenerators Using
           Hydrotreating                                              11

1-4        FCC's with CO Boilers                                      13

1-5        FCC Expansions                                             14

1-6        Simulated Crude Slates and FCC Feed Sulfur Content         17

1-7        Gas Oil Yields and Sulfur Content                          18

II-l       Range of Conditions for Fluid Catalytic Cracking           23

II-2       Typical Product Distributions                              24

II-3       Current Emission Rate from Typical FCC Unit                25

III-l      Pilot Plant Performance Data for Exxon Once-Through
           Sodium System    '                                          42

III-2      Status of Exxon Jet Ejector Venturi Scrubber Applications
           to the FCC Systems                                         43

III-3      Desulfurization Processes                                  55

III-4      Achievable Emission Levels with Best Control Techniques    57

IV-1       States with FCC Capacity of 1% or Greater                  59

IV-2       State Air Regulations                                      60

IV-3       State Air Pollution Standard for SO  Emissions from
           FCC Regenerator                                            61

IV-4       SOX and Hydrocarbon Emission Reduction from Typical
           Plant Based on Achievable Control                          63

IV-5       State Regulations for Hydrocarbon Emissions                64

V-l        SOX Concentration in Flue Gas from FCC Regenerators
           for Different Crudes                                       70

-------
                           LIST OF TABLES cont.


Table
 No.                                                                 Page

V-2        SOX Concentration in Flue Gas from FCC Regenerators        72

V-3        SOX Emissions from Recycle Oil                             73

V-A        SOX Concentration in Uncontrolled Emissions                75

V-5        Estimated SOX Emissions from FCC in 1978                   79

V-6        Estimated SOX Emissions from FCC in 1985                   80

V-7        Summary of Input/Output Variables for Model IV             86

VI-1       Estimated Cost of FCC Regenerator Modifications and
           New Equipment                                              90
                                    vi

-------
                              LIST OF FIGURES

Figure
 No.                                                                 Page

1-1         Petroleum Administration for Defense (PAD) Districts      10

II-l        Typical Fluid Catalytic Cracking Unit                     20

II-2        Feed Type and Coke Sulfur Level                           30

II-3        Sulfur Content of Coke                                    31

III-l       Jet Ejector Venturi Scrubber                              38

III-2       Schematic of Exxon FCC Jet Ejector Scrubbing System       39

III-3       Wellman-Lord, Inc., Sulfur Dioxide Recovery, Sodium
            System                                                    45

III-4       Shell Flue Gas Desulfurization Unit                       49

III-5       S02 Concentration in FCC Regenerator Off-Gas              53

V-l         Projected FCC Capacity Additions and Replacements
            Affected by NSPS                                          67

V-2         SOX Content of CO Boiler Stack Gas                        71

V-3         SOX Emissions Attributable to Sulfur in Recycle Oils      74

VI-1        Evolution of Fluid Cracking (Unit C, 1947-73)             88
                                    vii

-------
                             EXECUTIVE SUMMARY
PURPOSE AND SCOPE

     Section 111 of the Clean Air Act charges the Administrator of the
Environmental Protection Agency (EPA) with the responsibility of establish-
ing Federal standards of performance for new stationary sources which may
significantly contribute to air pollution.  These new source performance
standards (NSPS) are to reflect the degree of emission limitation achievable
through application of the best demonstrated control methods considering
cost.

     The Scope of this task order is twofold:  (1) to identify and present
all available data that would define the emission levels that can be
achieved with the most effective demonstrated control systems, and (2) to
estimate the emission reductions that would result through promulgation of
new source performance standards for Fluid Catalytic Cracking (FCC) units
used by the petroleum refining industry.  The results of this task will be
used as part of the EPA's assessment of numerous industries for the purpose
of establishing priorities for setting standards.

OVERVIEW OF FLUID CATALYTIC CRACKING (FCC)

     As of late 1975, the U.S. petroleum refining industry had an aggregate
FCC charge capacity of 5.67 million barrels per stream day (BPSD) fresh
feed and recycle, of which recycle accounted for 16%.  Some sort of hydro-
treating is currently performed on 6% of the fresh feed.  Our investigation
indicated that there is no accurate audit of the number of CO boilers
installed on FCC units.  A rough estimate is that 60-70% of the capacity
has them.

     Improved catalyst regeneration by High Temperature Regeneration (HTR)
is the lastest technology advancement being applied to FCC units.  Based
on interviews with industry representatives, it is estimated that HTR is
being applied to about 15% of the domestic FCC capacity.  This is expected
to increase over the 10 year period considered in this evaluation.

     In general, FCC capacity in the United States is not expected to show
significant growth over this period.  Growth rates of 0.5-1.5% per year
are projected through 1985.

POLLUTANTS CONSIDERED

     The pollutants assessed under this task include carbon monoxide,
particulates, NOX, SOX, and hydrocarbons, with emphasis on the last two.
NSPS currently exist for CO emissions from FCC units.  The level of control
was established to be consistent with that attainable with HTR, thereby
including this technology as a method of control.  The current control level
for CO has a logical basis and no modification seems warranted.
                                    viii

-------
     The net reduction in 1985 of NSPS for SOX is 49 to 85,000 tons/year
depending on whether the level of control is set at 300 or 500 ppm.  The
impact is small due to the relatively small growth rate of new capacity
and the low rate of replacement of existing facilities.  No net reduction
is expected from the application of NSPS for hydrocarbons since all new or
substantially modified FCC units installed after 1975 will likely utilize
either HTR or CO boiler technology.  Hence, the emissions with NSPS based
on best technology will be the same as the effective emissions without
NSPS.

APPLICABILITY OF NSPS TO MODIFIED SOURCES

     One of the issues addressed in this assessment was whether a FCC
modification such as revamping for HTR would constitute a major recon-
struction as defined in the Federal Register.  One of our industry contacts
noted that the wording of the regulation can be interpreted in such a way
as to imply that the affected facility is only the regenerator not the
entire FCC unit.*  The key point is whether the cost of revamping a FCC
regenerator is more than 50% of the cost of a new regenerator or FCC unit
depending on how one interprets the law.

     The results of cursory review of typical costs for FCC regenerator
modifications and new equipment are presented in Table 2.  The cost of
revamping the regenerator is approximately 15% of the cost of a new facil-
ity (regenerator only).  Even if the revamp costs 2 to 3 times the figure
shown, it would not be classified as a major reconstruction, even with a
narrow interpretation of the regulations.  Consequently, NSPS will not
generally apply to reconstructed FCC units based on current practice.
*Informal discussions with the EPA indicates that their interpretation en-
 compasses the entire FCC unit, since the catalyst regenerator is an
 integral part of the system.
                                   xi

-------
                               TABLE 2

                  ESTIMATED COST OF FCC REGENERATOR
                   MODIFICATIONS AND NEW EQUIPMENT

                      Capacity - 45,000 Bbl/Day


                                               New          Modified

Capital Investment1 (1975)                           $106

   New FCC Unit                               21.6

   Regenerator                                 8.6            1.2

   Modified - Percent of New                   —              14
^Materials and labor.
Source:  Hydrocarbon Processing, 1974 Refining Handbook, and Exxon
         R&E.
                                 xii

-------
                    I.  THE PETROLEUM REFINING INDUSTRY
     The stated purpose of this study is to estimate the atmospheric emis-
sion reduction of potential New Source Performance Standards (NSPS) for
SOX and hydrocarbon emissions from fluid catalytic cracker (FCC) regenerators.
In order to do this it is necessary to first identify the FCC population and
some of the factors which bear on the emissions.  In this chapter we present
this information.

A.   EXISTING FLUID CATALYTIC CRACKING UNITS

     As of December 1975, the U.S. petroleum refining industry consisted
of 256 refineries with a total crude capacity of 15,687,000 barrels per
stream day (BPSD).^  One hundred and forty-three of these facilities have
fluid catalytic cracking units (FCCU), which are the focus of this study.
The aggregate FCC charge capacity is 5,675,100 BPSD9 fresh feed plus recycle.
Table 1-1 shows the location of refineries with FCC units (by state and
city), the operating company, total FCC capacity of the refinery (some re-
fineries have multiple FCC units), and the type of catalytic cracking tech-
nology used.

     Table 1-2 shows the geographical distribution of FCC capacity by state
and Petroleum Administration for Defense (PAD) District and the percentage
of total U.S. capacity by PAD.  (See Figure 1-1 for PAD Districts.)  PAD's
II and III together contain almost three-quarters of the total capacity,
with PAD IV containing only 4%.  Another useful way of breaking down the
FCC population is by Air Quality Control Region (AQCR).  A map of these
AQCR regions and a table showing the relationship between AQCR and PAD dis-
tricts may be found at the beginning of Appendix C.

     With regard to individual FCC capacity, the average (total capacity/
number of units) is approximately 50,000 BPSD.  However, seventy-five per-
cent of all FCC units are smaller than the average capacity so a typical
facility would be less than 50,000 BPSD.

B.   GENERAL OPERATING FACTORS

     Total capacity consists of both fresh feed and recycle.  According to
The Oil and Gas Journal figures, 930,190 BPSD, or 16% of the total, is re-
cycle.  FCC recycle ratios are strongly dependent on the type of catalyst
used, but have generally been declining as a result of the improved activity
afforded by zeolite catalysts.  Recycle rates in the range of 5-15% of
fresh feed are common for operations utilizing zeolitic catalyst.  A 5%
recycle is about the practical minimum since some cycle oil is required to
return the catalyst that is entrained into the fractionator with the reac-
tor effluent.

     Six percent of total FCC fresh feed is hydrotreated.  The distribution
of feed hydrotreating by PAD district is shown in Table 1-3.
                                    —1—

-------
                                     TABLE 1-1
                LOCATION OF U.S. FLUID CATALYTIC CRACKING CAPACITY

                                 (as of 12/31/75)
      Location
Arkansas

  El Dorado

California

  Carson
  Benicia
  Santa Fe Springs
  Torranee
  Martinez
  Santa Fe Springs
  Martinez
  Wilmington
  El Segundo

  Richmond

  Wilmington
  Bakersfield
  Los Angeles
Colorado

  Denver
  Commerce City

Delaware

  Delaware City

Hawaii

  Barbers Point


Illinois
   Operating Company
Lions Oil Co.
Atlantic Richfield Co.
Exxon Co.
Gulf Oil Co.
Mobil Oil Corp.
Phillips Petroleum Co.
Powerine Oil Co.
Shell Oil Co.
Shell Oil Co.
Standard Oil Co. of
  California
Standard Oil Co. of
  California
Texaco, Inc.
Toscopetro Corp.
Union Oil Co. of
  California
Continental Oil Co.
The Refinery Corp.
Getty Oil Co., Inc.
Standard Oil Co. of
  California
Process
 Type
 Fluid
 Fluid
 Fluid
Fresh
 Feed
(BPSD)
15,000
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Thermofor
Fluid
Fluid
Fluid
57,000
45,000
13,500
56,000
47,000
11,000
46,000
35,000
43,500
43,500
28, 000 *
12,000
45,000
15,000
75,000
62,000
14,100
Fresh Feed
 + Recycle
  (BPSD)
  18,000
                        65,000
                        58,000
                        13,800
                        56,0002
                        47.0001'2
                        12,000
                        86,000
                        40,000

                        54,500*

                        54,500l
                        28.0001'2
                        12.0003

                        52,000
                        16,000
                         7,900
  77,000
  23,000
Wood River
Blue Island
Amoco Oil Co.
Clark Oil & Refining
Corp.
Fluid
Fluid
38,000
24,000
42,000
25,000
!BPCD; BPSD not recorded.
2Recycle not recorded.
3No recycle.

-------
                                 TABLE 1-1 cont'd
                LOCATION OF U.S. FLUID CATALYTIC  CRACKING CAPACITY
    Location
Illinois cont'd

  Hartford

  Robinson
  Joliet
  Wood River
  Lawrenceville
  Lockport
  Lemont
Indiana

  Whiting
  E. Chicago
  Mount Vernon
  Indianapolis
Kansas

  El Dorado
  Arkansas City
  Coffeyville
  Phillipsburg
  Wichita
  Augusta
  McPherson

  Shallow Water

  Kansas City
  El Dorado

Kentucky
  Catlettsburg

Louisiana
  Lake Charles
  Lake Charles
(as of 12/31/75)

Operating Company

Clark Oil & Refining
Corp.
Marathon Oil Co.
Mobil Oil Co.
Shell Oil Co.
Texaco , Inc .
Texaco , Inc .
Union Oil Co. of
California
Amoco Oil Co.
Atlantic Richfield Co.
Indiana Farm Bureau
Cooperative
Association, Inc.
Rock Island Refining
Corp.
American Petrofina, Inc.
Apco Oil Corp.
CRA Inc.
CRA Inc.
Derby Refining Co.
Mobil Oil Corp.
National Cooperative
Refinery Association
North American Petroleum
Corp.
Phillips Petroleum Corp.
Skelly Oil Co.
Ashland Petroleum Co .
Cities Service Oil Co.
Continental Oil
Process
Type


Fluid
Fluid
Fluid
Fluid
Fluid
Fluid

Fluid
Fluid
Fluid


Fluid

Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Thermofor

Fluid

Thermofor
Fluid
Fluid
Fluid
Fluid
Thermofor
Fresh
Feed
(BPSD)

26,000
36,500
85,000
94,000
31, 000 l
30, 000 l

58,000
123,000
48,000


6,000

16,000
11,000
9,400
14,500
8,000
10,800
21,500

20,000

5,500
32,000
31,000
54,000
125,000
27,000
Fresh Feed
 + Recycle
  (BPSD)
  27,000
  44,500
  85,0002
  94,0002
  31.0001'2
  30.0001'2

  66,000
 130,000
  50,000
   6,000'

  16,000'
  11,500
  10,200
  16,000
   8,600
  12,500
  23,500

  21,000

   5.5002
  48,000
  48,000
  55,000
 145,000
  32,000

-------
                                 TABLE 1-1 cont'd
                LOCATION OF U.S. FLUID CATALYTIC CRACKING CAPACITY
(as of 12/31/75)

Location

Louisiana cont'd
Baton Rouge
Metaire

Belle Chasse
Alliance
Refinery
Meraux
Norco
Chalmette
Convent
Michigan
Bay City

Detroit
Alma
Minnesota
Wrenshall
Pine Bend
St. Paul Park

Operating Company


Exxon Co.
Good Hope Refineries,
Inc.


Gulf Oil Co.
Murphy Oil Corp.
Shell Oil Co.
Tenneco Oil Co.
Texaco Inc.

Bay Re fining-Dow
Chemical, USA
Marathon Oil Co.
Total Leonard Inc.

Continental Oil Co.
Koch Refining Co.
Northwestern Refining
Process
Type


Fluid

Fluid


Fluid
Fluid
Fluid
Fluid
Fluid


Thermofor
Fluid
Fluid

Fluid
Fluid

Fresh
Feed
(BPSD)

169,000

8,500


78,000
10,500
100,000
22,000
70, 000 1


6,000
21,500
12,000

9,500
41,000

Mississippi
  Purvis
  Pascagoula
Missouri
  Sugar Creek

Montana

  Laurel
  Billings
  Billings
  Great Falls
                       Co., Division of
                       Ashland                   Fluid     21,000
Ameralda-Hess Corp.
Standard Oil Co. of
  Kentucky
Amoco Oil Co.
Thermofor   14,500

  Fluid     56,000


  Fluid     41,000
Ceney                       Fluid     11,500
Continental Oil Co.         Fluid     14,000
Exxon Co.                   Fluid     19,000
Phillips Petroleum Co.      Fluid      1,800
                                                                      Fresh Feed
                                                                       + Recycle
                                                                        (BPSD)
                                                                       169,000s

                                                                         8,5002
                                                                        80,300
                                                                        11,000
                                                                       102,000
                                                                        22,0002
                                                                        70.0001' 2
                                                                         8,000
                                                                        25,400
                                                                        13,500
                                                                        10,000
                                                                        42,000
22,500


14,5002

58,000


53,000
                         14,500
                         21,000
                         34,000
                          3,000

-------
                                 TABLE 1-1 cont'd

                LOCATION OF U.S. FLUID CATALYTIC CRACKING CAPACITY
     Location
Nebraska
  Scottsbluff

New Jersey
  Perth Amboy
  Linden
  Paulsboro
  Westville

New Mexico
  Artesia
  Ciniza

New York
  Tonawanda
  Buffalo

North Dakota
  Mandan
Ohio
   Canton
   Cleves
   Toledo
   Lima

   Toledo

   Toledo


 Oklahoma

   Cyril
   Enid
   Ponca City
   Wynnewood
   Gushing

   Okmulgee
            (as of 12/31/75)


   Operating Company
CRA, Inc.
Chevron Oil Co.
Exxon Co.
Mobil Oil Corp.
Texaco Inc.
Navajo Refining Co.
Shell Oil Co.
Ashland Petroleum Co.
Mobil Oil Corp.
Amoco Oil Co.
Ashland Petroleum Co.
Gulf Oil Co.
Gulf Oil Co.
Standard Oil Co. of
  Ohio
Standard Oil Co. of
  Ohio
Sun Oil Co. of
  Pennsylvania
 Apco  Oil  Corp.
 Champlin  Petroleum Co.
 Continental Oil Co.
 Kerr-McGee Corp.
 Midland Cooperatives
   Inc.
 OKC Refining Inc.
 Process
  Type
  Fluid
Fresh
 Feed
"(BPSD)
 2,400
Houdriflow  30,000
  Fluid    130,000
Thermofor   25,000
  Fluid     40,000l
Thermofor
  Fluid
  Fluid
 5,200
 7,200
  Fluid     22,000
Thermofor   19,000
 23,000
Fresh Feed
 + Recycle
  (BPSD)
   2,900
              38,000
             150,000
              25,0003
              40,0002
   5.2002
  10,800
               22.0002
               25,000
  34,000
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Thermofor
24,460
18,000
19,800
37,700
52,500
50,000
6,700
19,000
44,000
11,500
7,000
8,000
25,200
27,000
21,800
45,500
71,500
57,500
8,375
19,300
44, 000 2
13,500
10,000
10,000

-------
                 TABLE 1-1 cont'd
LOCATION OF U.S. FLUID CATALYTIC CRACKING CAPACITY
(as of 12/31/75)

Location

Oklahoma cont'd
Duncan
Tulsa
West Tulsa
Ardmore
Pennsylvania
Marcus Hook
Philadelphia
Marcus Hook
Warren
Tennessee
Memphis
Texas
Mount Pleasant
Port Arthur

Texas City
Houston
Corpus Christi
Houston

El Paso
Corpus Christi

Big Spring

Houston

Sunray

Bay town
Port Arthur
Tyler
Texas City
Beaumont

Borger
Sweeny

Operating Company


Sun Oil Co.
Sun Oil Co.
Texaco Inc.
Vickers Petroleum Corp.

BP Oil Corp.
Gulf Oil Corp.
Sun Oil Co.
United Refining Co.

Delta Refining Co.

American Petroleum Inc.
American Petrofina,
Inc.
Amoco Oil Co.
Atlantic Richfield Co.
Champlin Petroleum Co.
Charter International
Oil Co.
Chevron Oil Co.
Coastal States Petro-
chemical Co.
Cosden Oil & Chemical
Co.
Crown Central Petroleum
Corp.
Diamond Shamrock
Oil & Gas Co.
Exxon Co.
Gulf Oil Co.
La Gloria Gilt Gas Co.
Marathon Oil Co.
Mobil Oil Co.

Phillips Petroleum Co.
Phillips Petroleum Co.
Process
Type


Fluid
Fluid
Fluid
Fluid

Fluid
Fluid
Fluid
Fluid

Thermo for

Thermofor

Fluid
Fluid
Fluid
Fluid

Fluid
Fluid

Fluid

Fluid

Fluid
Thermofor
Houdrif low
Fluid
Fluid
Fluid
Fluid
Fluid
Thermofor
Fluid
Fluid
Fresh
Feed
(BPSD)

25,000
30,000
18, 000 l
20,000

40,000
80,000
75,000
10,000

13,500

9,600

30,000
157,000
69,000
13,000

24,000-
22,000

19,000

24,000

43,000
11,500
11,500
125,000
120,000
10,000
28,500
84,000
24,000
55,000
30,000
                                                      Fresh Feed
                                                       + Recycle
                                                        (BPSD)
                                                        35,500
                                                        31,400
                                                        18.0001' 2
                                                        21,000
                                                        41,600
                                                        86,500
                                                        85,000
                                                        10,200
                                                        13,500s
                                                        11,800

                                                        32,000
                                                       204,000
                                                        74,000
                                                        13,105

                                                        29,000
                                                        30,000

                                                        19,600

                                                        25,000

                                                        52,000
                                                        13,500
                                                        13,500
                                                       146,000
                                                       126,000
                                                        15,000
                                                        33,000
                                                        84,0002
                                                        24.0002
                                                        70,000
                                                        35,000

-------
                 TABLE 1-1 cont'd
LOCATION OF U.S. FLUID CATALYTIC CRACKING CAPACITY
(as of 12/31/75)

Location

Texas cont'd
Deer Park
Odessa
Corpus Christ!

Corpus Christi
Amarillo
El Paso
Port Arthur
Texas City
Nederland

Fort Worth
Utah
Salt Lake City
Salt Lake City

North Salt Lake
Roosevelt
Woods Cross
Virginia
Yorktown
Washington
Ferndale
Anacortes
Anacortes
Wisconsin
Superior
Wyoming
Casper
Cheyenne
Cody
Casper

Sinclair
Newcastle
Casper


Operating Company


Shell Oil Co.
Shell Oil Co.
Southwestern Refining
Co . , Inc .
Suntide Refining Co.
Texaco , Inc .
Texaco , Inc .
Texaco , Inc .
Texas City Refining Inc.
Union Oil Co. of
California
Winston Refining Co.

Amoco Oil Co.
Chevron Oil Co.

Husky Oil Co.
Plateau, Inc.
Phillips Petroleum Co.

Amoco Oil Co.

Mobil Oil Corp.
Shell Oil Co.
Texaco , Inc .

Murphy Oil Corp.

Amoco Oil Co.
Husky Oil Co.
Husky Oil Co.
Little America Refining
Co.
Pasco Inc.
Tesoro Petroleum Corp.
Texaco, Inc.
7
Process
Type


Fluid
Fluid

Fluid
Fluid
Fluid
Fluid
Fluid
Fluid

Fluid
Fluid

Fluid
Fluid
Houdriflow
Thermo for
Fluid
Thermofor

Fluid

Thermofor
Fluid
Fluid

Fluid

Fluid
Fluid
Fluid

Thermofor
Fluid
Thermofor
Fluid

Fresh
Feed
(BPSD)

70,000
10,500

9,500
20,000
8.0001
7.0001
135, OOO1
27,000

40,000
3,400

18,000
10,000
8,000
4,400
5,200
8,000

27,000

25,500
36,000
27, OOO1

9,700

9,500
10,000
3,300

6,500
17,700
4,000
7, OOO1

                                                      Fresh Feed
                                                       + Recycle
                                                        (BPSD)
                                                        70,OOO2
                                                        15,500

                                                        12,000
                                                        26,500
                                                         8,OOO1*2
                                                         7,OOO1' 2
                                                       135,OOO1* 2
                                                        28,000

                                                        44,000
                                                         6,000
                                                        22,000
                                                        11,000
                                                        13,000
                                                         6,900
                                                         5.2002
                                                        10,500
                                                        32,000
                                                        27,500
                                                        53,000
                                                        27,OOO1' 2
                                                        10,700
                                                        11,000
                                                        12,500
                                                         4,300

                                                        10,500
                                                        19,900
                                                         7,000
                                                         7,OOO1'2

-------
                                TABLE 1-2
              U.S. FLUID CATALYTIC CRACKING CAPACITY BY PAD

                        (Fresh Feed and Recycle)
  PAD
District
           Delaware
           New Jersey
           New York
           Pennsylvania
           Virginia

           Total
Fresh Feed
  (BPSD)

 ( 62,000)
 (229,444)
 ( 41,000)
 (206,000)
 ( 27,000)

 (565,444)
Fresh Feed
 &  Recycle
  (BPSD)

   77,000
  269,444
   47,000
  224,300
   32,000
Percent
of Total
               649,744
   12
           Illinois
           Indiana
           Kansas
           Kentucky
           Michigan
           Minnesota
           Missouri
           Nebraska
           North Dakota
           Ohio
           Oklahoma
           Tennessee
           Wisconsin

           Total


           Arkansas
           Louisiana
           Mississippi
           New Mexico
           Texas

           Total


           Colorado
           Montana
           Utah
           Wyoming
(429,277)
(193,000)
(163,700)
( 54,000)
( 39,500)
( 71,500)
( 41,000)
( 2,400)
( 23,000)
(202,460)
(191,200)
( 13,500)
( 9,700)
(1,434,237)
( 15,000)
(617,778)
( 70,500)
( 12,400)
(1,257,166)
(1,972,844)
( 22,500)
( 46,300)
( 53,600)
( 58,778)
523,277
203,800
206,450
55,000
46,900
74,500
53,000
2,900
34,000
248,500
231,675
13,500
10,700

18,000
677,728
76,850
17,560
1,527,571

23,900
72,500
70,160
74,078
                          1,704,202
                            30
                          2,317,709
                            41
           Total
  (181,178)
               240,638

-------
                            Table 1-2 cont'd




              U.S. FLUID CATALYTIC CRACKING CAPACITY BY PAD
PAD
District

5

California
Hawaii
Washington
Total
Fresh Feed
(BPSD)
(485,611)
( 14,100)
( 91,500)
(591^2111
Fresh Feed
& Recycle
(BPSD)
621,211
23,000
118,600

Percent
of Total

762t811 13
U.S. TOTAL
(4,744,914)
5,675,104
Source:  The Oil and Gas Journal, March 29, 1976.

-------
PETROLEUM ADMINISTRATION FOR DEFENSE (PAD) DISTRICTS
   rV r~~~--
   i 9 »***
   r
                      FIGURE 1-1

-------
                                TABLE 1-3

              CAPACITY OF REFINERIES WITH FCC REGENERATORS
PAD 1
New York
Pennsylvania
Total
PAD 2
Illinois
Michigan
Minnesota
Ohio
Total
PAD 3
Louisiana
Mississippi
Texas
Total
PAD 4
Wyoming
Total
PAD 5
California
Washington
Total
GRAND TOTAL
USING HYDROTREATING
Fresh Feed
(BPSD)
20,000
40,000
27,000
12,500
20,000
22,500
41,000
23,000
70,000*
16,000
58,000
8,500

60,000
82,000
134,000
16,000
76,500
368,500
                                                            Approximate
                                                            % of Total
                                                                10
                                                                10
                                                                  , **
**6,000 BPD listed as cat cracker and cycle stock feed pretreatment
  and middle distillate.

**Percent of total U.S. capacity.

Source:  The Oil & Gas Journal.  March 29, 1976.
                                    11

-------
     In undertaking this study we also Investigated what fraction of the
population have carbon monoxide (CO) boilers.  Data available from the
National Emission Data System (see Table 1-4) shows 31 facilities with 28%
of fresh feed capacity have CO boilers.  However, conversations with state
and local environmental control agencies and others (documented in
Appendix C) have led us to believe that the figure is closer to 60-70% of
total capacity.

     The current technology advance being applied to FCC units is the appli-
cation of high temperature regeneration (HTR) or in-situ CO combustion.
The application of this technology improved catalyst regeneration and re-
duces CO emissions.  Based on interviews with industry representatives
(Appendix C), it is estimated that about 15% of U.S. catalytic cracking
capacity is based on HTR.

     FCC units represent a large investment and are normally operated at
full capacity.  Run periods for FCC units typically last three to four
years, after which the unit is shut down for repairs and maintenance.  On
the average, the repair and maintenance period is about one month.  Thus,
the normal fractional utilization of existing capacity can be assumed at
97%.  Run lengths of six years without shutdown have been achieved.

C.   GROWTH PROJECTIONS

     Projecting growth rates for the petroleum industry is extremely dif-
ficult at present due to the ambiguity of government environmental and
energy policies.  In addition, new refining facilities require enormous
capital investments which are being deferred until future requirements are
better defined.

     The growth of FCC capacity is particularly affected by these uncertain-
ties because it is a key process for production of both motor and middle
distillate (heating and diesel) fuels.  The FCC process has a certain
amount of inherent flexibility which permits optimization of the yields
to match product demands for these fuels.  Therefore, FCC capacity is tied
directly to the growth projections for gasoline and middle distillate.

     Current U.S. fluid catalytic cracking capacity (fresh feed) is 4,745
K BPSD based on the latest Oil and Gas Journal Annual Refining Survey.9
Planned additions to capacity, all of which will be completed by 1976,
amount to 81.5 K BPSD or about 1.7% of 1975 capacity (see Table 1-5).
There are no domestic FCC construction projects beyond 1976 indicated in
the recent Oil and Gas Journal worldwide construction survey.12

     As indicated earlier, gasoline demand is an important factor effecting
expansion of FCC capacity.  Cantwell11 of DuPont recently presented gaso-
line demand projections at the API meeting in Los Angeles.  Two scenarios
were presented:

     Scenario 1:  Gasoline growth of 1.5% per year through 1980, peaking
     in 1980 at 8% above 1975 level and then declining at 1.6% per year
     back to 1975 level in 1985.

                                    12

-------
                                 TABLE 1-4

                          FCC'S WITH CO BOILERS*
     State^                   Operator

1    California     Atlantic Richfield
     California     Atlantic Richfield
2    California     Gulf Oil Corp.
3    California     Mobil Oil Corp.
4    California     Powerline Oil Co.
5    California     Standard Oil of California
6    California     Texaco, Inc.
7    California     Union Oil Co. of California
8    California     Shell Oil Co.
9    California     Exxon Co., USA
10   Delaware       Getty Oil Co.
11   Hawaii         Standard Oil of California
12   Illinois       Texaco, Inc.

13   Kansas         Petrofina Co.
14   Kansas         Skelly Oil Co.
15   Kansas         Apco Oil Corp.
16   Kansas         National Co-op Refinery
                      Association
17   Kansas         CRA, Inc.
18   Kansas         CRA, Inc.
19   Kansas         Phillips Petroleum
20   Kentucky       Ashland Oil Co.
21   Louisiana      Exxon Co., USA
22   Louisiana      Shell Oil Co.
23   Montana        Exxon Co., USA
24   New Jersey     Exxon Co., USA
25   Ohio           Ashland Petroleum Co.
26   Oklahoma       Vickers Petroleum Corp.
27   Texas          Shell Oil Co.
28   Texas          Exxon Co., USA
29   Utah           Caribiou (Phillips)
30   Utah           Chevron Oil Co.
31   Utah           Amoco
           Location

Carson  Pt 17
Carson  Pt 18
Santa Fe Springs
Torrance
Santa Fe Springs
El Segundo
Wilmington
Los Angeles
Wilmington (Carson)
Benicia
Delaware City
Barbus Point
Lawrenceville (St. Lawrence
  Fid.)
El Dorado
El Dorado
Arkansas City

McPherson
Coffeyville
Philllpsburg
Kansas City
Catlettsburg (Leach)
Baton Rouge
Norco
Billings
Linden
Canton
Ardmore
Deer Park
Baytown
Woods Cross
Salt Lake City
Salt Lake City
*Partial listing.
Source;  National Emission Data System Point Source Listing (latest update
         of data ranges from 1970-1974).
                                    13

-------
Company and Location
Champlin Petroleum Co.
  Corpus Christ!, Texas
   TABLE 1-5

FCC EXPANSIONS

 (Fresh Feed)

          Net
        Capacity
        Addition
         (BPSD)


         41,000
 Capacity
  Before
Expansion
  (BPSD)
  10,000
   Year
    of
Completion
   1976
Charter International Oil
  Hous ton, Texas
         26,000
  24,000
   1976
Exxon Co., USA
  Baytown, Texas
          5,000
 124,000
   1976
Good Hope Refineries
  Louisiana
          9,500*
  15,000
   1976
     Total Additions
         81,500
*16,000 BPSD expansion to 24,500 BPSD total capacity.  Net addition is
 24,500 - 15,000 - 9,500 BPSD.
Source:  The Oil and Gas Journal, Construction Survey, 1976.
                                    14

-------
     Scenario 2:  Gasoline growth of 2.3% per year peaking in 1981 at 15%
     above 1975 level and then declining at 1.6% per year to 8% above 1975
     level by 1985.

     Scenario 1 assumes that small cars will increase their market share
from 50 to 60% by 1980, that vehicle miles per car will remain constant
and that fuel economy will improve as required by EPA.  Scenario 2 presumes
consumer resistance to small cars and only partial achievement of fuel
economy goals.

     The average compound growth rate over the ten year period is about
3/4% per year for scenario 2.  The growth rate of FCC capacity over this
same period is projected at 1/2% per year by one of the process licensors
(see Appendix C).  A catalyst supplier we contacted forecasts FCC growth
at about 2% per year through 1980 and probably tapering off after that.

     Using this information, we adopted the following growth projection as
a basis for the impact calculations.


                     PROJECTED GROWTH OF FCC CAPACITY

                        1.5% per year through 1980
                         0% per year through 1985


This is an average compound growth over 10 years of about 3/4%.  The initial
growth is also about the same as from 1975 to 1976.

     It is most probable that some of this additional capacity will be ob-
tained through modification of existing units to remove bottlenecks or to
improve yields by applying HTR or feed hydrotreating.  However, since the
overall growth projections are very modest, we have assumed for the purpose
of calculating the impact of NSPS, that the additional capacity would come
from new units.  The impact under this assumption will represent a maximum
benefit case.

     For the purpose applying the Model IV calculation procedure, the fol-
lowing growth or conversion to HTR was assumed.

                                             HTR, %
                                               of
               Year                    Fresh Feed Capacity

               1975                            15
               1978                            25

               1980                            33
               1985                            60
                                    15

-------
D.   SULFUR CONTENT OF FCC FEED

     In order to calculate present and future SOX emissions nationwide, It
was necessary to estimate the average sulfur content of current and pro-
jected FCC feeds.  This was done by PAD district, excluding PAD District
IV, where the refining capacity is insignificant.  PAD District II is sub-
divided into large and small refinery operations for better characteriza-
tion of the industry.  Likewise PAD III is characterized by the two large
refining states contained in the district.

     Typical crude slates for the years 1977, 1980, and 1985 have been
established previously for EPA by ADL.13  The general objective in selecting
crude slates for each district was to simulate as closely as possible the
average mixture of crudes which would be run in each refining area in a
given year.  Specifically, the crude slates were chosen to simulate the
average domestic/foreign mix, the sulfur content, the API gravity, and
other key properties.  The crude slates are shown by PAD district in
Table 1-6.

     Also shown in Table 1-6 is the average sulfur content of FCC feeds in
future years.  These sulfur levels were estimated using the yield data in
Table 1-7 and assuming that the fresh catalytic cracker feed was composed of
both heavy gas oil and vacuum gas oil; i.e., everything in the boiling range
500-1050°F.  For determining the feed sulfur content, this is a good approx-
imation.  The sulfur levels in Table 1-6 do not take into consideration
hydrotreated feedstocks.  This is considered in Chapter V.

     The sulfur content of FCC feeds is shown to increase with time in PAD
Districts I and II due to increased use of foreign high sulfur crudes.  In
PAD V there is a slight decline in sulfur content after 1977 showing the
effect of Alaskan crude which is slightly lower in sulfur than crudes pre-
sently processed.  These trends were taken into account in the development
of emission baselines and uncontrolled emission levels which are discussed
in Chapter V.
                                    16

-------
                                         TABLE  1-6

                SIMULATED CRUDE  SLATES AND FCC  FEED  SULFUR CONTENT
Pad District
I East Coast




II Midwest - Small
(Okla., Kan., etc.)




Midwest - Large
(111., Ind., etc.)



Ill Louisiana Gulf

Texas Gulf




V West Coast





Crude
Type
L
T
N
A
V
AL
L
T
0
MC
A
AL
L
T
0
A
MC
T.
T
L
T
N
A
V
cu
cv
A
MC
I
ANS
Crude (vol. Z)
73
15.4
7.6
16.2
7.6
31.7
21.5
7.6
13.1
61.5
17.8
—
—
6.0
70.3
4.9
8.5
10.3
88.7
11.3
47.4
41.4
3.8
5.3
2.1
37.4
13.8
31.3
7.1
10.4
__
77
—
17.2
35.7
26.5
20.6
6.0
12.0
59.8
8.8
6.7
6.7
2.9
63.8
4.7
23.5
5.1
88.2
11.8
47.4
41.4
3.8
5.3
2.1
40.0
13.2°
33.4
3.4
10.0
—
80
—
18.2
40.7
21.5
19.6
2.9
10.9
58.0
—
14.1
14.1
60.1
4.6
35.3
—
88.2
11.8
47.4
41.4
3.8
5.3
2.1
40.0
13.2
—
—
—
46.8
85
—
19.2
45.7
16.5
18.6
10.0
55.8
—
17.1
17.1
55.1
4.4
40.5
—
88.2
11.8
47.4
41.4
3.8
5.3
2.1
40.0
13.2
—
—
—
46.8
Average Wt. Z S
in Catalytic Cracker Feed
73 77 80 85

0.82 1.13 1.16 1.19





0.47 0.53 0.60 0.64



1.42 1.56 1.69 1.70



0.40 0.41 0.41 0.41


1.54 1.54 1.54 1.54




1.31 1.35 1.09 1.09



LEGEND:  L  - Louisiana
        T  - West Texas Sour
        0  - Oklahoma
        CW - California Wilmington
CV  - California Ventura
ANS - Alaskan North Slope
N   « Nigerian Forcados
A   • Arabian Light
V  » Venezuelan Tia Juana
AL - Algerian Hassi Messaoud
MC " Mixed Canadian
I  • Indonesian Minas
                                             -17-

-------
                                                     TABLE 1-7
                                         GAS OIL YIELDS AND SULFUR CONTENT
00
         Crude

Louisiana
West Texas Sour
Oklahoma
California Wilmington
California Ventura
Alaskan North Slope
Nigerian Forcados
Arabian Light
Venezuelan Tia Juana
Algerian Hassi Messaoud
Mixed Canadian
Indonesian Minas
Yield of
Heavy Gas Oil
(500-650°F)
(Vol. %)
19.50
14.11
11.77
11.96
12.53
15.50
20.60
15.01
12.70
11.92
15.20
15.00
Yield of
Vacuum Gas Oil
(650-1050°F)
(Vol. %)
32.50
29.60
28.04
38.54
32.08
29.49
30.40
29.50
32.80
22.71
25.60
41.00
Wt. % S
in H.G.O.

0.0901
1.2187
0.1296
0.9124
1.1393
0.4547
0.2015
1.0807
0.6690
0.0756
0.4362
0.0361
Wt. % S
in V.G.O.

0.3221
1.8513
0.2327
1.3126
1.5411
1.1029
0.3125
2.3215
1.6292
0.2249
0.7121
0.0890
       Source:   Refinery Model  Study,  Appendix H,  "The Impact  of Lead Additive Regulations on the Petroleum
                Refining Industry,"  by Arthur D. Little,  Inc., EPA Contract 68-02-1332,  Task Order 7,
                December 1975.

-------
                    II.  SOURCES AND TYPES OF EMISSIONS
     There are six different licensors of FCC technology.  However, since
catalytic cracking has been an important commercial process for over thirty
years, the various processes have evolved toward each other.  Design dif-
ferences among the processes do not affect the character of the emissions
from catalyst regeneration.  Therefore, FCC technology with conventional
regeneration is treated for purposes of this study as though there were
only one licensor.

A.   PROCESS DESCRIPTION

     Figure II-l is a schematic flow diagram of a typical fluid catalytic
cracking process.  The major pieces of equipment are the regenerator, the
riser cracker, the catalyst separator, and the fractionator.  Fresh feed
is mixed with recycle from the fractionator before entering the riser
cracker, where the bulk of the cracking reactions occur.  The riser cracker
is designed for plug flow (i.e., the catalyst is entrained in the oil vapors)
to maximize the catalyst activity and selectivity, since the amount of
carbon-on-catalyst is the smallest at the bottom.  Hence, the term "fluid"
catalytic cracking is a misnomer in that the cracking no longer takes
place within a fluid bed.  The substitution of riser cracking for fluid
beds was initiated by the introduction of the highly active zeolite cata-
lysts, which reduce the required cracking time to a few seconds.

     The catalyst separator is designed to disengage the catalyst and
cracked products as rapidly as possible to avoid subsequent recracking.
Internal cyclones remove some of the catalyst fines from the product vapors.
The product vapors travel overhead to the fractionator.  Essentially all
the entrained catalyst fines exit in the column bottoms stream, which ulti-
mately goes to a settler.  A catalyst sludge is removed from the settler
and is mixed with the liquid recycle for reinjection into the riser reactor.

     In the separator, the catalyst, now containing deposited coke, falls
through a steam stripping section, where residual heavy oil is removed.
The catalyst activity is restored by burning off the carbon in the regen-
erator, with air.  The combustion gases exit the regeneration through in-
ternal cyclones which recover catalyst fines.  The regenerator flue gas
may then be routed to an electrostatic precipitator and/or a CO boiler.

     The only source of airborne emissions from the process is the flue
gas, which may contain CO, SOo, S0-j, hydrocarbons, NOX and particulates.
Hence, the remainder of the discussion will be limited to catalyst regen-
eration.

B.   REGENERATION

     There are basically two types of regeneration:  (a) traditional and
(b) complete CO combustion, otherwise known as high-termpeature regenera-
tion  (HTR).  HTR is a relatively recent innovation, and, consequently it
                                    19

-------
                                                                                                               Gas and Gasoline
ts>
O
                       Cyclones
             Riser
            Reactor
                                                              	|
          	I
                                         FIGURE II-l   TYPICAL FLUID  CATALYTIC  CRACKING UNIT

-------
is used in a minority of FCC units.  HTR offers significant advantages over
traditional regeneration, most of them directly affecting the emissions.
These advantages are as follows:

     1.  More complete catalyst regeneration.  Both the catalyst activity
         and selectivity are improved; less coke is produced per unit of
         feed, which reduces the SOX emissions, because the sulfur on the
         catalyst going to the regenerator is related to the carbon on the
         the catalyst.

     2.  Lower catalyst inventory.  The complete regeneration increases
         the catalyst activity, which means that a lower catalyst-to-oil
         ratio is possible.  In turn, the unit's capacity or severity can
         be increased if bottlenecks are removed from the rest of the pro-
         cess.

     3.  Better heat recovery within the regenerator.  The major oxidation
         reactions are as follows:

                                             Heat Released
              Reaction                      (Btu/lb carbon)

              C + 1/2 02  ->•  CO                  4,440

              CO + 1/2 02  -»•  C02               10,160
              C + 02  -»•  C02                    14,160


         Traditional regenerators do not allow complete CO oxidation, and,
         hence, a large quantity of heat is either lost or must be recovered
         in a CO boiler.

     4.  Low CO emissions.  It is possible to meet the 500 ppm CO NSPS
         without a CO boiler.

     Due to these advantages, some "traditional" units have been revamped
for high-temperature regeneration.  Generally, revamping requires the re-
placement of cyclones, the plenum chamber, cyclone diplegs, the regenera-
tor grid and seals, and the catalyst overflow weir, all of which must be
type 304 stainless steel rather than carbon steel in order to withstand the
higher temperatures, resulting from HTR.

     Essentially complete catalyst regeneration at lower temperatures can
be achieved through the use of CO oxidation promoter catalysts.  These
catalysts are identical in terms of cracking performance to their non-
promoter counterparts, however, they contain metals which catalyze the
regenerator reaction

                             CO + 1/2  ->  C02
                                     21

-------
Hence, a lower regenerator temperature can be used to meet the CO NSPS
without upgrading the metallurgy of the regenerator internals for HTR.  The
promoter catalysts are considerably more expensive, however, which means
that there is a tradeoff between capital expenditure and increased catalyst
cost.  The SOX emissions from a process using the promoter catalyst would
be greater than the HTR case at the same cracking conditions.  Regeneration
of the catalyst is not quite as effective at the lower temperature; hence,
the selectivity of the catalyst is slightly poorer in that more coke is
produced.

C.   CONDITIONS

     Table II-l lists the ranges of conditions for the FCC process includ-
ing both traditional and high-temperature regeneration.  Processes using
a promoter catalyst would have a regeneration temperature typically in the
range of 1160-1200°F.

Yields

     Expected yields vary considerably with feedstock for the three generic
categories of feedstocks—aromatic, naphthenic, and paraffinic as indicated
in Table II-2.  Generally, the highest conversions (conversion is defined
as 100 minus percent cycle oil) are obtained from paraffins and the lowest
from aromatics, which are more refractory.  Most of the difference in yield
structure between paraffins and naphthanes lies in the higher gas make for
the former.14

     The concentration of aromatic molecules determines the coke yield,
and, hence, is important with regard to emissions.  Hydrotreating the feed
can convert some of the aromatics to naphthenes, thereby reducing the coke
make and increasing the conversion.  In fact, it has been shown that hydro-
treating catalyst cracker feed to increase the conversion can be economi-
cally attractive in certain cases.15

D.   EMISSIONS FROM A TYPICAL FCC UNIT

     The major source of airborne emissions (exclusive of fugitive emissions)
in an FCC unit is the regenerator flue gas.  As shown in Figure II-l, this
gas may be routed through an electrostatic precipitator and/or a CO boiler;
the order of the precipitator and boiler may also be reversed.  In addition,
there may be a heat recovery unit downstream of the regenerator (not shown).

     In Table II-3, emissions for a typical FCC unit are summarized.  Point
A lies between the electrostatic precipitator and the CO boiler; typical
emissions for both types of regeneration are indicated.  Point B lies
downstream of the CO boiler; only the traditional regeneration case is
considered at this point, since it is assumed that high-temperature regen-
eration is sufficient for meeting the CO NSPS.  It should be noted, however,
that the use of HTR does not necessarily preclude the use of CO boilers.
Some refiners (Exxon, for instance) prefer to use medium-temperature regen-
eration to obtain the benefits from improved catalyst selectivity without
                                    22

-------
                               TABLE II-l

           RANGE OF CONDITIONS FOR FLUID CATALYTIC CRACKING
Reactor

  Temperature, °F

  Pressure, psig

  Catalyst/Oil Ratio (wt.)

  Riser Velocity, ft/sec

  Coke Content of Spent Catalyst, wt. %

  Recycle Ratio, % Recycle to Fresh Feed
 885-1025

   9-40

   3-20

  15-70

0.25-6

   0-100
Regenerator


  Temperature, °F - Traditional Regeneration
                  - HTR

  Pressure, psig

  Coke Content of Regenerated Catalyst, wt. "A
1000-1100
1100-1350

   9-40

0.05-1.0
                                  23

-------
                                              TABLE II-2
                                     TYPICAL PRODUCT DISTRIBUTIONS
reedstock Description
Conversion, vol. % 1
Yields, vol. %
     Gasoline, C5-430°F. TBP ep
     Butane-butene
          isobutane
          n-butane
          b u t e n e s
     Propane-propylcne
          propane
          propylcnc
     Light cat gas oil
     Decanted oil
     Total
     Coke, wt. %
     C~ and lighter, wt. %
Aromatic
70
54.2
16.8
5.9
1.4
9-5
7.5
2.0
5.5
20.0
10.0
108.5
6.3
3.0
0.4
Naphthenic
85
70.0
19.0
7.3
1.9
9.8
8.5
2.4
6.1
10.0
5.0
112.5
5.4
2.8
0.2
Paraffinic
93
73.0 '
22.5
8.0
2.5
12.0
12.0
3.3
8.7
5.0
2.0
114.5
4.8
2.5
0.1
f.ourro:  The Oil and Gas Journal - October 30, 1972
 Conversion defined as 100 minus percent cycle oil.

-------
                                                      TABLE II-3
                                      CURRENT EMISSION RATE FROM TYPICAL FCC UNIT

                                                Capacity - 50,000 BPSD

Significant
Emission Sources
POINT A3
Traditional
Regeneration
High-Temperature
Regeneration
POINT B4
Traditional
Regeneration

Gas Volume
SCFM
87,050
(4.3% H20)
92,500
(3.4% H2)
110,950
110,950

(Wet
Part.
gr/dscf
0.055
0.043
0.043
0.043
Gas
Basis
sox1
1,360
1,070
1,070
1,070
Composition -
Except
NO 2
X
200
200
200
200
ppm

for Particulates)
CO
114,000
500
10
10
HC
500
10
10
10
°2
Nil
9,800
10,000
10,000
Pollutant
Emission

Rate-lb/hr
Part. SO
X
39.4 1,200
32.8 1,000
39.4 1,200
39.4 1,200
NO CO
X
127 44,000
134.4 205.0
162 4.9
162 4.9
HC
110.3
2.3
2.8
2.8
ISJ
Ul
     See Text for Discussion of Data Sources and Assumptions
     Expressed as S02
     Expressed as
     3lies between electrostatic precipitator and CO boiler
           downstream of CO boiler

-------
requiring the replacement of the regenerator internals .  In this regard ,
the cases in Table II-3 can be considered extremes, and the emissions for
the typical medium-temperature case would be somewhere in-between.

     The sources of the emission data in Table II-3 are discussed below.

1.   Gas Volume

     For the calculation of gas volume, a typical coke composition was
assumed.  The appropriate chemical equations for the regenerator reactions
were written and from these the gas volume was calculated for each case.
A sample calculation is given in Appendix A.  The assumptions were as
follows :

     •  The coke composition is 96% C, 3% H, 1% S

     •  The coke yields were as follows : 16

        Traditional regeneration - 6% by weight of feed
        High-temperature regeneration - 5.0% by weight of feed

     •  The C02/CO mol ratio leaving the regenerator is 1.0 for tra-
        ditional regeneration; only C02 leaves the regenerator for the
        HTR case.

     •  For the CO boiler and for HTR, the excess air was assumed to
        be 5%.10

     •  It was assumed that the CO boiler did not use auxiliary fuel.

2 .   Particulates

     Particulate emissions are restricted by an NSPS to 1 kg particulates
per 1000 kg coke burn-off so this was used as a typical emissions level.
To calculate the particulate emissions, it was assumed that the coke
yields were as shown above.  The gas volume calculated for part 1 were
used to determine the concentrations .

3.   S0y
     SOX emissions vary widely, depending primarily upon the cracking con-
ditions and  the feedstock quality  (discussed in Chapter I) .  Since essen-
tially all the carbon is removed from the catalyst by regeneration, all
the sulfur contained in the coke is converted to SOX and is emitted with
the flue gas.  The typical SOX emissions shown in Table II-3 were calcu-
lated as follows.  The average rate of SOX emissions (per barrel of
fresh feed)  in the United States for 1978 was estimated (see Chapter V)
and scaled up to 50,000 BPSD, the  result being 1,200 Ib/hr as indicated
for the two  traditional regeneration points in Table II-3.  For the same
cracking conditions, HTR reduces the coke production — typically to 5.0 wt. %
                                   26

-------
of fresh feed for HTR from 6.0% for traditional regeneration.16  Since the
SOX emissions are roughly proportional to the coke make, the emission rate
for the HTR case was scaled down by 5.0/6.0 ratio.  The gas concentrations
were calculated assuming 1) gas volumes as shown In Table II-3, 2) SOX as
SC>2, and 3) regenerator off-gas molecular weight, 29.

     It has been found by several groups that a portion of the SOX is pres-
ent as 803, apparently formed by the reaction equals

                           S02 + 1/2 02  -»•  S03

There is disagreement as to the percentage of SO  as SO.; estimates have
ranged from 0.1-60%.  Amoco indicated that concentrations of 803 well in
excess of the calculated equilibrium value for the reaction above under the
regenerator conditions have been observed.  However, this phenomenon is
presently unexplalnable.  Exxon stated that the 803/802 ratio is generally
0.03 from their units.  Clearly, more data are needed since a high 803
concentration could have a large impact on the operation of certain scrubber
systems.

4.   N0_
     Very few data are available on NOX emissions.  The Monsanto Report
summarized data from previous studies and the range of emissions downstream
of a traditional regenerator was 8-394 ppm.  Downstream of the CO boiler
it was 0-500 ppm.  Interviews with Exxon and Amoco provided the following
estimates:

                           After HTR             After CO Boiler
                           (ppm NO )                (ppm NO )

           Exxon              <200                    <200

           Amoco              5-20                     <50

It was assumed that a typical NOX emission was 200 ppm for each case.  How-
ever, there is obviously a paucity of useful data.  In addition, the data
for emissions downstream of the CO boiler are probably biased upward due
to the common practice of burning supplemental fuel in the boiler.

     In general, one would expect the NOX emissions from HTR to be lower
than those from a CO boiler due to the lower temperature; this is indicated
by the Amoco figures.

5.   Carbon Monoxide

     Carbon monoxide emissions are now restricted to 500 ppm by a NSPS.
The standard was set to allow HTR units to meet the standard without a CO
boiler.  Hence, 500 ppm was chosen as a typical emission level from high-
temperature regeneration.
                                     27

-------
     The CO level between a traditional regenerator and the CO boiler was
calculated by stoichiometry using the assumptions in part 1.

     The emission level downstream of the CO boiler was estimated from
actual stack sampling data.17  For five cases, the CO concentrations were
0, 0, 5, 10, and 25 ppm; the average is approximately 10 ppm, so this
figure was assumed for Table 11-3.

6.   Hydrocarbons

     The level of hydrocarbon emissions is very low.  Interviews with Amoco
and Exxon produced the following estimates:


                              After HTR          After CO Boiler

           Exxon          <5 (non-methane)              —
           Amoco                 <10                   <10

UOP reasoned that hydrocarbon emissions are practically nil because the
combustion reactions in a boiler or in a regenerator are rate-limited by
the combustion of carbon monoxide—not by the combustion of hydrocarbons.

     P.O.M.'s or P.N.A's

     Importance is attached to the polynuclear aromatics in the hydrocar-
bons due to their potential carcinogenic effect.  The P.N.A. present in
significant quantity in FCC flue gas is Benzo-A-Pyrine (BAP).  The concen-
tration of BAP in the flue gas from conventional regenerator is 0.481 Ibs/
1,000 barrels.18  The BAP concentration is effectively reduced in the CO
bWiler.  The concentration of BAP in the flue gas from CO boilers is
0.031 x 10~3 lb/1,000 barrels.  It is expected that the BAP concentration
from the HTR units is comparable to the BAP concentration in the CO boiler
flue gas; however, there is no data to establish this as fact.

7.   Oxygen

     Downstream of a traditional regenerator, the oxygen concentration is
insignificant because less-than-stoichiometric air is supplied.  For the
two remaining cases, the oxygen in the stack is set entirely by the excess
air supplied, and is about 10,000 ppm in each case here (5% excess air).

E.   FACTORS AFFECTING SO  EMISSIONS

     This discussion is confined to SOX emissions, since other species are
of lesser importance in terms of the impact of NSPS (see Chapter III).

     The two most important parameters affecting SOX emissions are the coke
yield and the feed sulfur content.  The coke yield is important since all
the sulfur in the coke is ultimately discharged as SOX.  The feed sulfur
content is significant since there appears to be a correlation between
                                    28

-------
feed sulfur content and coke sulfur content, though the relationship is not
well understood.

     Since high-temperature regeneration (HTR) produces a more selective
catalyst, coke make is 20-30% lower than that for traditional regeneration.
Hence, HTR units create smaller SOX emissions.  Other than the type of
regeneration, the parameters affecting coke yield are feed type (discussed
earlier)  and  the severity of operating conditions.  Since heavier feeds
produce more coke, coke production is expected to increase in this country
as heavier foreign crudes are substituted for domestic crudes.

     There are two published articles19* 20 on the subject of sulfur distri-
bution in catalytic cracking; they are in general disagreement In their
attempts to correlate coke sulfur content with the feed sulfur content.
Data from Amoco are presented as a plot of the ratio (wt. Z sulfur in coke)/
(wt. % sulfur in feed) versus conversion (see Figure II-2).  The result is
a series of parallel lines with small slope; the sensitivity to conversion
is smaller than expected.  The ratio of sulfur in coke to sulfur in feed
varies from about 0.5-0.8 for virgin (unhydrotreated) feeds and from
approximately 0.3-0.6 for the same feedstocks, but hydrotreated.  It is
surprising that the ratio would be lower for the hydrotreated feedstocks
since hydrotreating does not remove much of the multi-ring thiophenes, which
are the species expected to remain in the coke.  The Amoco authors concluded
that there is "no clear correlating parameter using the sulfur-type analysis.'
However, it was stated that this may be due to the uncertainty created by
the large number of unanalyzed sulfur compounds.

     A separate article by Gulf20 (see Figure II-3) indicates that the
above ratio is 2-3 for low-sulfur feeds—obtained by hdyrotreating higher-
sulfur feeds—and 1-2 for the unhydrotreated stocks.  These data are directly
opposite from the Amoco data.  The Gulf authors imply that there jjs a cor-
relation between feed sulfur type and coke sulfur content.  The Gulf data
are indirectly supported by a Chevron paper21 .

     Although one would expect a correlation between feed sulfur content and
coke sulfur content, the relationship has not been adequately developed.
Nevertheless, it can be said that, the higher the feed sulfur, the higher
the coke sulfur and, consequently, the higher the SOX emissions.  As the
heavier and more sour foreign crudes are substituted for domestic crudes,
increases in SOX emissions are expected nationwide.
                                     29

-------
1
s
4"
**
\
8 04
.S ° '

»« *•*
£
0






MU-C.nlln.nl wM< b.11
K.il I.I.I bh)b billing
H,*illHf 2
Zeolite B (ololysl,
conversion increasing with
talalyil/oil ond tnnptrotuii
SO 55 60 65 70
V
















75 W
J
FIGURE II-2   FEED TYPE AND COKE SULFUR LEVEL
Source:  Wollaston, E.G., "Sulfur Distribution
         in FCU Products," The Oil and Gas Journal,
         August 2, 1971.
                      30

-------


f
so
4.0
1.0
7.0
s«
$
.
— 10
M OS
* 0.4
03
0.2
O.I
V

legend
O W Trx gos oil
A Kgnci.t slr,(ki
& Kuwi.l G 0 . U*0 Linn)
O So. In. gm oil
O Cobindo 
-------
            III.  BEST APPLICABLE SYSTEMS OF EMISSION REDUCTION
     The emissions from FCC regenerators and the factors that affect emis-
sions were described in the previous chapter.  The pollutants include par-
ticulates, CO, SO , NO  and hydrocarbons.  In this chapter, the control
technology to reduce the pollutant emissions is discussed.  This informa-
tion may be used in the establishment of New Source Performance Standards
(NSPS).

     At present, NSPS are implemented for particulates and CO; therefore,
control technology for these pollutants is not discussed here.  A discussion
of NOX control technology is also omitted, since SO  and hydrocarbons are
of primary concern.  Pollution control technology for SOX  and hydrocarbons
is well-developed and the majority of this chapter is devoted to describing
control systems for these pollutants.  The control status of all the pol-
lutants is briefly discussed first.

A.  CONTROL STATUS

1.  Particulates

     The existing NSPS for particulate emissions is 1 kg/1,000 kg of coke
burned, or about 0.038 gr/scfd.  This control level can be achieved during
the FCC on-stream period by well-maintained cyclones followed by an electro-
static precipitator.  Hence, the present control level is consistent with
best available technology.

     The current control level may be slightly beyond the capability of
low-energy wet scrubbers.  It is necessary to use a medium- or high-energy
scrubber to meet NSPS for particulates.

2.  Carbon Monoxide

     The existing NSPS for carbon monoxide in FCC regenerator flue gases
is 500 vppm.  This level was established based on data provided by Amoco
that demonstrated that high-temperature regeneration could reduce the CO
emission below 500 ppm.  Carbon monoxide emissions after CO boilers are
generally less than 50 ppm17 and many new units without HTR would likely
include CO boilers to increase the total energy efficiency of the refinery.
However, purchasers of new FCC units would also consider the benefits of
yield improvements and lower capital costs (no CO boiler) associated with
HTR.  If the NSPS on CO were reduced below 500 ppm, this latter option of
HTR would be eliminated due to non-compliance.  A rational basis was used
to set the existing NSPS on CO and the logic still applies.

3.  Oxides of Nitrogen

     There is no NSPS for NO  emissions from FCC regenerators.  NOX emis-
sions are generally less than 200 ppm after the CO boiler and in the flue
                                   32

-------
gas from HTR, according to information obtained from interviews with in-
dustry representatives.

                               NO  EMISSION

               Source           Ex. HTR        Ex. CO Boiler

              Exxon R&E        <200 ppm          <200 ppm

              Amoco            5-20 ppm           <50 ppm

It appears that in some cases the concentrations are relatively low.

     Even if the concentration was of potential concern, effective control
technology for NO  is still being developed.  The enforcement of an NSPS
for this pollutant would, therefore, have to await commercialization of
control technology.

4.  Hydrocarbons

     The hydrocarbon emissions depend on the type of regeneration (high-or-
low-temperature).  Also, CO boilers effectively reduce or even eliminate
the hydrocarbon emissions.  The hydrocarbon emissions in the FCC flue gas
at various locations are given as below.

                                             Hydrocarbon Emissions
                     Source                    in Flue Gas, ppm

        Conventional Regenerator                     <500

        CO Boiler                                     <10

        High Temperature Regenerator, HTR             <10

     The NSPS only apply to new and modified sources.  At present, a NSPS
to control hydrocarbon emissions does not exist.  It is expected that new
or modified FCC's will utilize HTR or have CO boilers, therefore, a NSPS
would have a minimal effect on the control of hydrocarbon emissions from
FCC's.  In effect, the NSPS for CO has simultaneously controlled the emis-
sion of hydrocarbons to relatively low levels.  The effect of a possible
NSPS standard for hydrocarbons is shown in Chapter V.

     The control technologies to reduce hydrocarbon emissions are HTR or
use of a CO boiler.  Since HTR is discussed in the previous chapter and
the petroleum industry is familiar with CO boilers, control technology for
hydrocarbon reduction is not further discussed in this chapter.

5.  Oxides of Sulfur

     At present, there is no NSPS for SO  emissions from FCC regenerators
The sulfur content of the feed to FCC's nas been increasing.  It is esti-
mated that the SOX concentration in the flue gas will increase from 885 ppm

                                    33

-------
mated that the SO  concentration In the flue gas will increase from 885 ppm
in 1978 to 905 ppm in 1988 (national averages, see Chapter V).  The control
technology for SO  reduction is reliable and includes various processes to
suit the individual situation.  The processes suited for application to
FCC regenerators are described in detail in the latter part of this chapter.

6.  Other Pollutants

     Other pollutants from FCC regenerators include HCN and NH,.  The con-
centration of both of these pollutants is 200 ppm each in flue gas from
traditional regenerators and 10 ppm each in flue gas after CO boilers.
These chemicals are destrueted at high temperature.  The control technology
for reduction of these pollutants is the same as the control technology
for reduction of hydrocarbons.  Therefore, the same logic for application
of NSPS applies.

     It should be mentioned that the reduction of HCN and NH- in HTR or
CO boilers will increase the concentration of NO  emissions.  For example,
the NO  emissions in flue gas from conventional regenerators is <10 ppm;
however, the NO  emissions in flue gas after CO boilers is 100-200 ppm.
(Some NO  in CO ooilers forms from nitrogen in the air supply and nitrogen
in the o¥l feed to the CO boiler.)

B.	SO  EMISSION CONTROL
	x	

     Reduction of SO  emissions can be accomplished by a variety of methods
including:

     •  Use of low-sulfur fuel as FCC feedstock,

     •  Removal of sulfur from the FCC feedstock, and

     •  Removal of SO  from FCC regenerator flue gas.
                     X

     Removal of sulfur from the FCC feedstock can be accomplished by direct
hydrodesulfurization since FCC feed is low in asphaltene materials.  Switch-
ing to low-sulfur feeds is the easiest option to implement and requires
minimal capital investment.  However, it is not a universal solution be-
cause there is not enough low-sulfur crude available to satisfy the poten-
tial demand.  Hence, in most cases, desulfurization of FCC feedstock or
desulfurization of regenerator flue gas must be employed to satisfy the
sulfur emission requirements.

     Various sulfur removal processes have been, or are being, developed
in each of the sulfur removal methods mentioned above.  In the remainder
of this chapter, we discuss both SO  removal methods as they have potential
application in the refinery industry.
                                   34

-------
C.  FLUE GAS DESULFURIZATION PROCESSES

     Flue gas desulfurization (FGD) is the removal of sulfur compounds
after combustion of the fuel.  Fuel combustion is associated with FCC
operation, since the coke deposited on the catalyst is removed by com-
bustion with air.  Sulfur present in the coke appears in the flue gas
in the form of SO- and SO.,.  The flue gas from the FCC regenerator may
contain S03 up to 50% of the total SO ;10 although the actual SO./SO-
ratio is uncertain.  The high concentration of SO, (compared to other
combustion processes) may be due to the presence of metallic compounds
such as vanadium which may act as a catalyst in the formation of S0~.

     The advantages of FGD systems (compared to desulfurization of FCC
feed) are flexibility of feed supply and lower energy consumption.

     Flue gas desulfurization processes may be classified according to
the final form of the sulfur removed from the flue gas as follows:

     •  Elemental sulfur

     •  Direct sulfuric acid

     •  Waste salts, and

     •  Concentrated S02.

     Only three processes, Exxon's jet ejector scrubber (a waste salt
process), the Wellman-Lord process (concentrated S02 process), and the
UOP process (concentrated S02 process), have actually been applied to
refineries (the Claus plant or boilers) and only the Exxon scrubber has
been used with a fluid catalytic cracker.  Because of this, we have
concentrated the evaluation in this report on these three processes appli-
cable to the refinery industry; however, other processes are mentioned
briefly.

1.  Elemental Sulfur Processes

     These processes produce elemental sulfur qr hydrogen sulfide
directly, with no intermediate production of concentrated S0».  Hydrogen
sulfide can be converted directly to elemental sulfur using commercially
available technology (the Claus process).  EPA is currently reviewing
proposals based upon some of these processes for funding of a 100-megawatt
coal-fired demonstration plant for a sulfur recovery process.  Processes
in this category include:

     •  Charcoal absorption - Westvaco, Foster-Wheeler

     •  Citrate process - Arthur G. McKee

     •  Potassium Thio sulfate system - Conoco Coal Development
                                   35

-------
     •  Sodium phosphate process - Stauffer Chemical/Chemico

     •  Dry carbonate process - Atomics International

     •  Ammonia process - Catalytic

These processes are in the earlier stages of development, and cannot be
expected to have any important commercial impact until about the mid-1980's.

2.   Direct Acid Processes

     In these processes, sulfuric acid is produced directly with no in-
termediate concentrated SO- gas stream.  A more dilute acid is produced
than the normal commercial 98% acid grade.  Operations involving this
approach have been unsuccessful to date and the outlook is not promising.

3.   Waste Salt Processes

     a.  General

     These processes produce varying types of sulfide and sulfate salts
which must be disposed of in an environmentally acceptable manner.  They
account for over 90% of the systems currently committed for SO  removal
and can be expected to sustain this strong position in applications which
will become operational over the next ten years.  Of these waste salt
or "throw-away" processes, a very large fraction involve some version of
lime or limestone slurry scrubbing to produce a solid waste calcium
sulfite/sulfate for disposal.  Most of these systems are applied to large
units where the gas flow rates are 300,000 scfm and higher.  The processes
based on lime/limestone suffer from problems of scaling of the scrubber
due to the presence of the insoluble salts in the scrubber liquor.
Consequently, process availability has been low.  Also, disposal of sludge
from FGD processes in refineries may be a problem.

     Soluble salts, such as sodium salts, may be used to avoid the scaling
problems in the scrubber; however, sodium salts are expensive.  In some
processes sodium salts are regenerated using lime or limestone, which
results in the formation of waste cake.  If the sodium solution is not re-
generated (once-through), it is necessary to reduce the COD of the waste-
water before discharging the scrubber bleed.  In some cases it may also
be necessary to reduce the insoluble and soluble salts.

     In the once-through sodium system the spent liquor from the scrubber
is not regenerated.  If the particulates and SO  are removed simultaneously
in the scrubber, the problems present in the regeneration of the scrubber
solution are increased.  In a once-through sodium system, particulates and
SO  can be removed simultaneously without these major problems.
  X
     Various types of contacting devices can be used to remove particulates
and SO .  These devices include venturi, jet ejector, tray tower, packed
      X
                                   -36-

-------
tower, TCA scrubber, spray scrubber, etc.  The particulate collection
efficiency of the spray scrubber is very low.  Collection devices such
as tray scrubbers, packed scrubbers, and TCA scrubbers are likely to plug
up if particulate matter is present.  The remaining devices such as con-
ventional venturi scrubbers and jet ejector venturi scrubbers can be
used to remove particulates and SO  simultaneously.

     b.  Exxon Jet Ejector Liquid Scrubber Process Description

     The application of jet ejector liquid scrubber to FCC units for
simultaneous removal of SO  and particulates, marketed by Exxon, is
shown in Figure III-l.  The schematic of the FCC jet ejector scrubbing
system is shown in Figure III-2.

     The scrubber liquid passes through a spray nozzle which breaks the
liquid stream into droplets.  These droplets have size and velocity
characteristics suitable for contacting with the flue gas to provide
the scrubber efficiency.

     The dust-laden flue gas is drawn into the body of the scrubber by
the draft-inducing action of the liquid spray.  The gas is intermixed
with the scrubbing liquid and both enter the venturi section of the
scrubber.  In the venturi section, the liquid and gas enter an area of
intense turbulence.  Here, the liquid droplets strike and capture the
particles in the gas stream and the SOX in the flue gas is absorbed
in the liquid.  Also, within this venturi^section momentum transfer to
the gas occurs creating a pressure build-up across the unit which will
move the flue gas through the system from the inlet of the scrubber to
the outlet.

     After passing through the venturi section, the mixture of gas and
droplets is sent to a separator where the clean gas is separated from
the contacted water and discharged to the atmosphere, after reheating.

     A conventional venturi scrubber operates in much the same way as
the jet ejector scrubber; except that momentum transfer takes place
from the gas phase to the liquid phase.  The liquid is broken up into
small droplets due to the high gas velocity in the venturi scrubber.
Power plants and steam plants have long used wet gas scrubbing of flue
gases to remove SO  and particulates using conventional venturi scrubbers
with associated stack fans to achieve pressure balance.  The difference
between the two is that in the conventional venturi scrubber the energy
is provided by a fan on the gas side, whereas in the jet ejector scrubber
energy is provided by a pump on the liquid side.  In the conventional
venturi scrubber, it is also necessary to have a liquid pump to provide
the flow of liquid to the venturi.

     In the scrubber, absorption of S02 takes place with the following
chemical reactions:
                                  -37-

-------
Spray
Nozzle
   Venturi
                 Scrubbing
                  Liquid
                                                Dirty Gts
                                           To JLiQUlD
Source:   Exxon R&E.
           Figure  III-l   Jet Ejector Venturi Scrubber
                                -38-

-------
1   ro  w^S
                            Figure III-2    Schematic  of  Exxon  FCC  Jet  Ejector  Scrubbing  System

-------
       Absorption:  2 NaOH + S02 -»• Na2S03 + H20           (1)

                    Na2S03 + S02 + 1/2 H20 -> 2 NaHS03     (2)

       Oxidation:   Na2S03 + 1/2 02 -»• Na2S04              (3)

     The purge stream from the scrubber will contain soluble salts of sul-
fites and sulfates.  The concentration of the soluble salts is generally
below 5%.  If the particulate collection and the SO  absorption are
carried out simultaneously in the wet scrubber, theXscrubber bleed will
contain insoluble salts in addition to soluble salts.  It will be
necessary to treat the scrubber bleed before it can be discharged.  The
water treatment should include removal of insoluble salts, reduction of
COD, and in some cases reduction of soluble salts.

Technical Considerations

     The venturi scrubbers are in general very flexible and changes in
the gas flow rate, temperature of the gases, and the humidity of the gas
have very little effect on the collection efficiency.  Furthermore, the
outlet concentration of the SO  in the flue gas can be easily controlled
by adjusting the pH of the scrubber solution.

     The collection efficiency of the scrubber can be changed by changing
the input energy to the scrubber.  This can be done by increasing the
pressure drop in the venturi for conventional venturi scrubbers and in-
creasing the pressure of the motivating fluid in the jet ejector scrubber.
Thus, the collection efficiency of the scrubber can be changed to meet
the future considerations.

     As discussed earlier, the scrubber may not be able to reduce the
outlet SO  concentration to the same level when the inlet SO  concentra-
tion is h¥gh.  This is because the jet ejector scrubber represents a single
stage co-current operation.  If more stages are required, provision should
be made for additional contact stages downstream of the ejector in the
form of a tray or a packed scrubber.  The gases coming into the tray or
packed scrubber will be particulate-free and, therefore, the problem of
scrubber plugging will not be present.

     The gases coming to the scrubber are cooled with spray water in
order to decrease the flue gas temperature.  The SO- in the flue gas will
condense and form mist at this temperature.  It will be difficult to capture
these submicronic mist particles since they will generally escape through
the venturi scrubber.  The SO- concentration in the Exxon FCC units has
been found to be 1-3% of SO-; thus, the problem of SO,, mist has not been
experienced in the Exxon facilities.

     In a once-through sodium system, the scrubbing solution is not re-
generated.  Particulate and SO  removal can therefore be conveniently
carried out in the same equipment which will reduce the system's initial
capital cost.


                                  -40-

-------
     If particulate removal Is carried out in the scrubber, the liquid
bleed from the scrubber will contain insoluble solids.  It is expected
that the handling of this slurry will increase erosion and solids deposi-
tion problems.  The liquid handling system should therefore have erosion-
resistant materials where slurry is contacted.

     The scrubber bleed will contain both soluble and insoluble solids,
making it necessary to treat the wastewater before it can be discharged.
The insoluble solids may be removed in the thickener/clarifier.  Treatment
of the clarified wastewater may be combined with treatment of other
refinery waste streams, which will lower the operating cost of pollution
control for FCC units.

     The caustic used in a once-through sodium scrubbing system represents
a significant operating cost.  The caustic consumption is directly pro-
portional to the inlet SO  concentration.  Thus, if the inlet SO  concen-
tration is increased from 500 ppm to 3,000 ppm, the caustic consumption
will increase sixfold.  The economics of the system will therefore vary
depending on the inlet SO  concentration.
                         X

Pollutant Removal Efficiency

     The pilot plant performance of the Exxon jet ejector scrubber is
shown in Table III-l.  The SO  removal efficiency was 95-99% with SO
inlet concentrations of 200-500 ppm.  Although the SO  concentration
in the inlet to the scrubber may be as high as 3,000 ppm for potential
applications, no data are available on the operation of the jet ejector
scrubber for inlet SO  concentrations higher than 500 ppm.
                     X
     The catalyst particulate concentration in the pilot plant inlet
gas was 0.1-0.3 gr/scf.  Under these conditions the collection efficiency
was 85-95%.  In general, the particulate concentrations to the scrubber
may vary from .02-1.5 gr/scf.  If the particle size distribution remains
the same, the collection efficiency will remain constant.  Thus, the out-
let grain loading will increase in direct proportion to the inlet grain
loading.

     The NO  removal efficiency of this scrubber is expected to be low.
           X

Status of Process

     The applications of Exxon jet ejector scrubber to FCC units are shown
in Table III-2.  Overall performance has been good.  There have been small
metal loss problems in the scrubbing liquid delivery system due to erosion
and corrosion, but solutions to these problems are being implemented.  The
service factors on the two Baytown scrubbers have been 94% and 93%, respec-
tively, since the beginning of 1975.  This service factor is lower than
for FCC in general, therefore, occasional by-passing of the scrubber would
be necessary to avoid outage of the FCC.  The SO  removal efficiency and the
particulate removal efficiency have been >90%.
                                  -41-

-------
                     TABLE III-l
            PILOT PLANT PERFORMANCE DATA
        FOR EXXON ONCE-THROUGH SODIUM SYSTEM
 Range of Pollutant           Collection Efficiency
Loadings to Scrubber          	Percent	

SO

    200-500 ppm                     95-99

Particulate

    0.1-0.3 gr/SCF                  85-95

Condensables

    0.1-0.2 gr/SCF                    <90
Source:  Exxon Research and Engineering
                          -42-

-------
                             TABLE III-2
      STATUS OF EXXON JET EJECTOR VENTURI SCRUBBER APPLICATIONS

                         TO THE FCC SYSTEMS
       Location

Baytown, Texas

Baytown, Texas

Baton Rouge, Louisiana

Bayway, New Jersey
Design Size
   ACFM

  300,000

  510,000

  860,000

  590,000
 Startup Date

March 1974

May 1975

February 1976

May 1976
Source:  Exxon Research and Engineering
                                  -43-

-------
4.  Concentrated SO  Processes

     These processes produce concentrated SO- gas streams from dilute
flue gas S0_.  Conventional, commercially proven technology is available
for conversion of concentrated SO- produced in these processes to either
liquid S0_, sulfuric acid, or elemental sulfur.  Liquid SO- is produced
by compression of the SO  stream after the gas from the process has been
concentrated and dried.

     The conventional technology involving reaction of H-S and SO- to
produce sulfur (Glaus process) can be used.  H-S is normally generated in
the refineries.  Thus, the processes generating SO- are attractive for ap-
plication to FCC regenerators.

     The processes generating SO- may be either wet or dry.  The leading
wet process is the Wellman-Lord process and the leading dry process is
the Shell/UOP process.  These are described in the following sections.

     a.  Wellman-Lord Process26

Process Description

     The Wellman-Lord process was developed by Davy Powergas, Inc., formerly
Wellman-Power Gas, Inc., of Lakeland, Florida.

     The flow scheme for the basic process is shown in Figure III-3.  The
S0_-rich gas is contacted countercurrently in the absorber by the sodium
sulfite solution and exits the absorber top stripped of SO .  The solution
leaving the bottom of the absorber, now rich in bisulfite, is discharged
to a surge tank and then pumped to a proprietary evaporator/crystallizer
in the regeneration section.

     Low pressure steam is used to heat the evaporator and drive off SO-
and water vapor.  The sodium sulfite precipitates as it forms and builds
a dense slurry of crystals.

     The gas stream leaving the evaporator is subjected to partial conden-
sation to remove the majority of the water vapor before the product S0_
is discharged from the process.  The condensate in the mixture with the
sulfite slurry stream withdrawn from the evaporator is used for re-dissolving
the slurry in the dissolving tank.  The sulfite lean solution is then
pumped to a surge tank and fed back to the absorber.

     Gas precleaning is recommended in the Wellman-Lord system.  If the
gases are cooled in the precleaning device, however, the SO., will be present
as fine mist particles.  Collection of these submicron particles is
difficult in a low energy scrubber, therefore, most of the particles will
escape the scrubber.

     The process is based on a sodium sulfite/bisulfite cycle.  In addition,
sodium sulfate (Na-SO,), which is nonregenerable in the normal process, is


                                  -44-

-------
                   Clean Gas 140 F
          Heat
        Recovery
Flue Gas
 HMXTF   ^-^  600°F


        H20
                                                         Evaporator
                        tvaporator  ,	»
                       Crystallizer J
                                                      Product
                                                   Sulfur Dioxide
                      Absorber
               Steam
€)
               Surge
               Tank
                                                                                         110°F
                                                                                 H2O
                                                                   Surge
                                                                    Tank
Dissolving
  Tank
            Figure III-3     Wellman-Lord, Inc., Sulfur Dioxide Recovery,  Sodium System

-------
formed in the absorber as a result of oxidation of sodium sulfite and
collection of sulfur trioxide.

     The sodium sulfate is controlled at a level of approximately 5 wt. %
in the absorber feed stream by use of a continuous purge from the system.
High concentration of SO. in the flue gas, if not removed in the gas pre-
treatment, would result In an increased concentration of Na-SO,  and pro-
portionally increased purge of the scrubbing liquor.

     A makeup of caustic is required to replace that lost in the purge
stream.  The caustic makeup solution reacts with the sodium bisulfite in
the absorber solution to form additional sodium sulfite.  Soda ash
(Na-CCO can also be used as the makeup source of sodium.

     The simple regeneration scheme of the patented Wellman-Lord process
relies on the favorable solubilities of the sodium system.  The bisulfite
form has almost twice the solubility of sulfite at the temperatures con-
sidered for the process.  Because of this it is possible to feed the absorber
with a saturated sulfite solution, or even a slurry, without any fear of
additional crystallization or scale forming despite considerable evapora-
tion of water.  This is because as S0_ absorption proceeds, the composition
of the solution is shifted in the direction of increasing solubility.

Technical Consideration

     The major concern with the Wellman-Lord system in this application is
sulfate formation.  The system may be unsuitable for application to the
FCC regenerator due to possible high sulfur trioxide levels in the gas
leaving the catalyst regenerator.  The S0»/S0- ratio in the catalyst re-
generator depends on process conditions and the presence of metallic
compounds such as vanadium which may act as oxidizing agents.  Therefore,
the ratio of S0»/S0_ in the flue gas will vary with changes in the process
conditions and changes in the oil feed.  The presence of SO. in high
concentrations will result in excessive sulfate formation.

     The Wellman-Lord process must avoid a buildup of contaminants in the
system liquor.  Possible contaminant sources include not only all by-
product formation in the system  (Na^SO,, Na-S-O-, S, etc.) but also all
soluble and insoluble contaminants picked up from the flue gas and process
makeup water.  However, the most significant of these is sulfate.  Since
it cannot be thermally regenerated sulfate must be purged from the system
in some type of bleed stream.  This purge also removes other contaminants
from the system; however, the purge rate will normally be determined by
the amount of sulfate formed.  The effect of this purge requirement will
be directly translated into process economics—in the capital investment
for the purge treatment facility and the operating costs associated with
the makeup sodium and purge disposal.  Because these costs can be substan-
tial and because process acceptability may be evaluated in terms of lost
sodium value and waste disposability, a considerable process development
effort has been expended in determining the various factors governing the
                                  -46-

-------
rates of sulfate formation and developing methods for minimizing both
sulfate formation and the attendant sodium losses in the required purge.

     The majority of the sulfate is formed in the absorption system both
from sulfite oxidation and from absorption of sulfur trioxide.  To some
degree the high solution concentrations used in the Wellman-Lord system
tend to reduce the sulfite oxidation because oxygen solubility is greatly
reduced at high solution concentrations.  However, this is not effective
against high SO, levels.  Work is being directed toward further reducing
oxidation by introducing anti-oxidants into the system.  The concentra-
tion of S0» present in the regenerator flue gas must be established before
the feasibility of using this process can be decided.

     The Wellman-Lord system is flexible in that it can handle wide varia-
tions in gas flow rate, temperature, humidity, and S02 concentration.
The exit SCL concentrations can be controlled by adjusting the pH of the
absorbing solution.

     The regeneration section and absorber are separated by surge tanks,
therefore, the regeneration section may be located away from the absorber
section.  If more than one absorber is used in the refinery, a common re-
generation system may be used.  These advantages may help reduce space
problems.

     The solution concentrations in the Wellman-Lord system are close to
saturation, thus, the flow rates in the system are small.  However, the
possibility of precipitation of salts is increased.

     The Wellman-Lord system generates SO^-rich gas.  This can be further
processed in a Claus-type facility at the refinery.

Pollution Problems

     The sulfate crystallizer in the Wellman-Lord system generates waste
solids.  If no impurities are present, it will contain 85% Na^SO,, 15%
Na_SO, crystals, and some process solution trapped in between the
crystals.  It will be necessary to dispose of these solids.

     The quantity of solids generated will depend on the amount of oxi-
dation of sodium sulfite to sodium sulfate and the absorption of 803
from the flue gas.  Presence in the flue gas of 1-2% SO, (percent of
total SO  present in the flue gas) and 2% 0  may result in 5-6% oxidation
of the sulfide solution.  This will generate dry waste solids representing
6-7% of the SO  absorbed.  The total waste solids will be dry solids
plus the liquor containing dissolved solids.

System Collection Efficiency

     The outlet SO,, concentration with the Wellman-Lord system may be as
low as in the once-through sodium system.  Concentrations in the range of
50 ppm are possible.
                                  -47-

-------
     The Wellman-Lord system has a tendency to accumulate impurities which
must be purged from the system.  Therefore, the particulate collection
must be carried out upstream of the scrubber.

     The system has poor collection efficiency for NO .   Efficiencies
for hydrocarbon and CO removal are negligible.

System Reliability

     The Wellman-Lord system is a fully-developed SO- control system.  It
has been commercially applied to nearly 25 flue gas sources, sulfuric
acid plants, Glaus plants, and utility boilers.  The system availability
is more than 95%.  It has not been applied to FCC units.

     b.  Shell Flue Gas Desulfurization Process (UOP)

     This process was developed by Shell International Research, The
Netherlands, and is presently licensed by Universal Oil Products, Des
Plaines, Illinois.

Process Description26

     The Shell Flue Gas Desulfurization (SFGD) process uses a dry acceptor
in a static packed bed to remove S02 from gaseous streams.  The CuO
acceptor contained in two or more identical reactors, is subjected to
successive stages of acceptance and regeneration at approximately the same
temperature.  The net effect is that S02, free of oxygen and particulates,
is obtained concomitant with the required degree of gas desulfurization.

     Figure III-4 is a simplified illustration of the SFGD process equip-
ment arrangement.  The S09-rich gas passes through the acceptance
reactor(s) for about 45-60 minutes, until the cumulative slip (break-
through) of S02 into the treated gas has reached a designated S02 concen-
tration.  The S02~rich gas stream is then switched to a reactor containing
regenerated acceptor, and the loaded acceptor is regenerated.  Gas from
the purging of the reactors between acceptance and regeneration is
treated in the accepting reactor.

     CuO is outstanding as an acceptor in this application in that it
readily forms sulfate with S02 in the presence of oxygen at, ideally,
about 700-750°F, and can be satisfactorily regenerated with reducing gas
to yield concentrated S0_ at about the same temperature.

Technical Considerations

     In general, adsorption systems have plugging problems; therefore,
the incoming flue gas must be free of particulate matter.  However, in
the Shell/UOP system the flue gas passes along the surface of the acceptor
packages and not through the acceptor material.  This prevents pressure-
drop buildup due to the deposition of particulate material present in
                                 -48-

-------
                   Treoted Flu* Go*
vo
I
                      Reducing
                        Go*
               0	
                 ••—0.0
                     T!   IT
                    Timing Device
 ii   i
1LLL
                           	
-------
the flue gas.  The system has been shown to operate on flue gas containing
particulate loading of 0.1 gr/scf with no evidence of plugging or fouling.

     In some FGD processes, oxidation or the presence of SO- creates prob-
lems during regeneration.  In the Shell/UOP process, however, the SO- does
adsorb; thus, the presence of SO- will not be a problem in this process.

     The particulate collection efficiency of the UOP adsorption system
is poor by design and the appropriate process location is downstream of
the ESP.  If sulfur trioxide were present in mist form, it would escape
through the adsorption reactor.  Since the adsorption is carried out at
high temperature (700-750°F), the SO-j will be present in the gas form
and will be adsorbed on the acceptor.

     Oxygen in the flue gas is used in the adsorption reaction and in
the reactions during regeneration.  The Shell/UOP process has been tested
on boiler flue gas which normally contains an oxygen concentration equal
to 4-6 percent.  SO  concentrations in the flue gas from a boiler are in
the same range as concentrations in the flue gas from a FCC regenerator.
The oxygen concentration in flue gas from the FCC regenerator system is
0.5 to 2 percent.  The effect of this lower oxygen concentration on the
SO  removal efficiency is not known.  The ratio of oxygen concentration
to SOX concentration in the FCC regenerator flue gas is 2 or higher
(minimum oxygen concentration of 0.5 percent oxygen and maximum SOX con-
centration of 2,500 ppm).  The theoretical oxygen to SOX mol ratio is
equal to 1.  Therefore, the effect of decreased oxygen concentration may
not be significant.

     Most of the FGD processes producing concentrated S02, concentrated
t^SO^, or elemental sulfur use reducing gas (natural gas, Ho, CO, etc.).
Reducing gases are generally in short supply and therefore the applicability
of these concentrated sulfur processes to FGD is limited.  In petroleum
refineries, however, reducing gas (H_) is generally available and there-
fore supply of hydrogen should not be a problem.

     As indicated in the process flow schematic, the reactor beds are
automatically cycled from acceptor mode to regeneration mode by the use
of sequential timers and motor-activated butterfly valves.  To prevent
leakage of reducing gas into the on-line reactor, special, large, tight
shut-off flue gas valves are required.  In addition, pressure surges
caused by reactor switching can influence operation of the upstream
process and aretherefore undesirable.  The effect of these surges is
minimized by the use of an "open bypass" which continuously recycles
a small quantity of treated flue gas.  The performance at existing facil-
ities has confirmed the operability of the sequential switching and by-
pass system.22

     The inlet temperature to the fixed bed reactor must be 700-750°F.
The flue gas temperature downstream of the particulate control device
(ESP) is normally in the range of 300-450°F.  Therefore, to obtain higher
flue gas temperature, less heat may be recovered in the CO boiler.  This
                                  -50-

-------
also requires a hot precipitator instead of a cold precipitator.  On the
other hand, the higher flue gas temperature may be obtained by using a
preheater.  The heat requirement in a preheater may be minimized by using
a heat exchanger to recover heat from the clean gas.  The preheater appli-
cation may be more suited for a new situation than for existing facilities
due to the arrangement of in-place equipment and space limitations.

     The flue gas from a wet system is saturated at the downstream of
the scrubber.  Therefore, it shows plume and sometimes results in acid
rain near the stack.  It is necessary to reheat the gases to avoid this
problem.  The problem of visible plume and acid rain is not present in
the dry systems.

Other Pollution Considerations

     The product from the Shell/UOP process is concentrated S02 gas.  The
gas can be sent to a Claus-type unit to generate sulfur.  Since the Shell/
UOP process is a dry system and therefore no liquid effluents are generated
in the process, it will be necessary to replace catalyst after 1-2 years
of operation.  This will create solid waste.  The quantity of solid waste
is very small.

Final Concentration

     The operation of the Shell/UOP system has demonstrated 90 percent
SO  removal efficiency when applied to boiler stacks.  It is expected
that proper operation of the system on FCC regenerator flue gas will have
the same efficiency.  By design, the system has poor collection efficiency
for particulate removal.  The SFGD system is capable of NO  removal by
addition of a suitable adsorbent according to UOP.  Performance data are
not known.

Status of Process

     The feasibility of the process was tested in a demonstration unit
erected in 1967 in the Shell refinery at Pernis, The Netherlands.  The
Pernis unit handled 600 scfm of flue gas containing 0.1 to 0.3 volume
percent S02>  A total of 20,000 hours of testing was accomplished to
establish the stable activity of an acceptor based on a reinforced
alumina support.  The estimated acceptor life is in excess of 8,000
cycles or 1-1/2 years.

     A commercial unit was started in August 1973 in Yokkaichi, Japan
to treat 80,000 scfm of oil-fired boiler flue gas containing a maximum
of 2,500 ppm S02 at the Showa Yokkaichi Sekiyu refinery. The S02 removal
efficiency is 90 percent and the hydrogen consumption is about 0.2 Ib
H~/lb sulfur removed.  The S02 produced in regeneration is totally
transferred to the Claus unit by the absorption/stripping section.

     The first installation of the SFGD process on a coal-fired boiler
in the United States is a 0.6 Mw pilot plant at Tampa Electric Company's


                                 -51-

-------
Big Bend Station which went on-stream in June 1974.  Bottled hydrogen is
used for regeneration and the S02 is returned to the stack.  The pilot
plant slip stream has been taken before and after the flue gas leaves a
cold electrostatic precipitator so that the CuO reactor has experienced
the full grain loading from the boiler.  The particulate entering the
reactor passes on through and there has been no evidence of plugging or
fouling.  The SC^ removal efficiency ranges from 75 to greater than 90
percent depending upon process parameters.

     The system has not been applied to desulfurize flue gas from FCC
regenerators.

D.  DESULFURIZATION OF FCC FEED

     The feed to the FCC may be desulfurized as an alternate to flue gas
desulfurization to meet SO  emission standards for FCC regenerators.
Desulfurization technology is very well known to petroleum refiners and
can be applied to almost any petroleum stock except certain heavy resids.
Typically, FCC feedstock (mainly vacuum gas oil) contains 0.51 to  .36
weight percent of sulfur.    However, vacuum gas oil obtained from
Louisiana, Nigeria Forcados, Oklahoma, Algerian-Hassi Messaoud, and
Indonesia, Minas charge stocks contain less than 0.5 percent sulfur.

     A portion of the sulfur in the FCC feedstock is deposited with the
coke on the catalyst.  The sulfur in the coke leaves as SOX during regen-
eration of the catalyst.  The sulfur concentration of the coke is 0.7-3
times the sulfur concentration in the FCC feed.  The sulfur in the coke
represents 2-30 percent of the total sulfur in the feed.20

     The range of sulfur concentration in coke and sulfur concentration
as a percentage of feed sulfur is large because the relationship is very
complex.  The thiophenic sulfur compounds in the feed have a tendency
to remain on catalyst with the coke.  The sulfur concentrations in the
coke as a percentage of feed to the FCC is high for feeds containing
thiophenic sulfur compounds.  Since hydrotreating removes essentially
all the non-thiophene sulfur compounds, the sulfur distribution for
hydrotreated feedstocks is significantly different than for virgin
feedstocks.

     Figure III-5 relates SOX concentration in the flue gas and coke
sulfur content.  R is the ratio of C02/C0 in the flue gas.  If the
concentration of CO in the flue gas is reduced to meet NSPS standards,
the value of R approaches <».  The concentration of CO in the flue gas
may be reduced by high temperature regeneration or using a CO boiler.
To achieve a 300-600 ppm SOX level, the coke sulfur content must be
0.36-0.72%.  Since the ratio of coke sulfur content to feed sulfur
content is 2-3, the feed sulfur must be reduced to 0.12-0.36% by hydro-
treating.

     The  technology for vacuum gas oil desulfurization is well-established.
Up to 97% of the feed sulfur can be removed, but typically, 90% removal
                                 -52-

-------
    3000
O-
Q_
(U
O
O
c

 cxi
O
to
 c
 O

 ro

 c
 
-------
is used as an economic guideline.  Thus, FCC feed desulfurization is a
possible control option to meet SO  emission standards.
                                  X

     There are several processes available for feed hydrodesulfurization.
These processes are shown in Table III-3.  All these processes are reliable
in operation and familiar to the refinery industry.

Process Description

     The FCC feedstock is mixed with hydrogen-rich gas, heated to treatment
temperature and passed through a reactor containing a fixed bed of desul-
furization catalyst.  The hot reactor effluent exchanges heat with the
incoming feed mixture and is cooled.  The recycle hydrogen is separated
from the liquid stream at high pressure, and recycled to the process to
minimize makeup gas hydrogen requirements.

     The liquid product stream is sent to a low pressure separator which
allows the removal of dissolved gases before the next stage, a stripper
column, where the product is freed from hydrogen sulfide and any light
ends.  Recycle hydrogen is also recovered from this low pressure separator.
Part of the heat recovered from the reactor effluent stream may be used
to supply heat to the product stripper reboiler or alternatively steam
can be used.

     All the processes are generally comparable and differ primarily in
the type of catalyst used.

Technical Consideration

     The hydrotreating (HT) of oil to remove sulfur is generally comparable
in expense to flue gas desulfurization and the final selection depends on
specific circumstances.  This is generally true for fuel combustion uses.
In FCC operation only part of the sulfur in the feed (2-30%) appears in the
FCC regenerator flue gas.  Thus, if feed desulfurization is carried out,
it will be necessary to remove considerably more sulfur than is present
in the FCC regenerator flue gas.  However, hydrotreating of FCC feed to
improve conversion, gasoline yield, and coke selectivity and run length
has been a recognized process tool.

     The hydrotreating will likely be applied on the basis of improved
yields or sulfur specifications in the FCC products rather than for S0x
standards on flue gas per se.

     Hydrotreating of feed requires large amounts of hydrogen and it is
assumed that a supply of hydrogen would be available in refineries—either
from reformer off-gas or from a hydrogen plant.
                                  -54-

-------
I
i-n
                   Process



       1)  Gulfining



       2)  Hydrofining



       3)  Hydrodesulfurization



       4)  Isomax
                                             TABLE III-3



                                      DESULFURIZATION PROCESSES






                                           Licensors



                                Gulf Research & Development



                                British Petroleum and Exxon



                                Institute Francaise du Petrole



                                Chevron Research or UOP
5)  Trickle Hydrosulfurization  Shell Development



6)  HGO-Unicracking             Union Oil Co. of California
       7)  Unifining





       8)  Gofining
                                Union Oil Co. of California

                                  and UOP



                                Exxon R&E
Operating    Total Distilled Capacity,

  Units      	BPSD	



    6       7,000-8,000 (unit capacity)



   84                  720,000



   83               >1,000,000



   25                  900,000


   82                1,050,000
       Source:  The Oil and Gas Journal.

-------
Other Pollution Problems

     The hydrotreating of the feed to FCC results in the formation of H?S.
These gases are converted to sulfur in Claus-type units.  There may be an
aqueous waste stream, if the oil hydrotreated contains nitrogen.  Also,
water is often injected upstream of heat exchangers to prevent plugging.

Sulfur Removal Efficiency

     The sulfur removal efficiency of hydrotreating may be as high as 97%.
A sulfur removal efficiency of 90% is typical of economic operation.  At
this efficiency, the sulfur content of vacuum gas oil feed may be reduced
to M).3% (maximum 3.3% sulfur in feed to hydrotreater).  Most of the
sulfur compounds in the hydrotreated feed will be thiophenic in nature and
a large portion of this sulfur will appear in the FCC regenerator flue gas.
With 0.3% sulfur in the HT feed, the SOX concentration in the flue gas
will be 400-500 ppm.  Thus, feed hydrotreating is not as effective a
method as FGD to reduce SO  in the FCC regenerator flue gas, if levels
below 400-500 ppm are required.

     The hydrotreating of feed can reduce NO  emissions.  However, the
extent of this reduction is not known.

Status of Hydrotreating

     The status of hydrotreating processes was presented in Table III-3.
These processes are commercially proven.

E.  ACHIEVABLE EMISSION LEVELS WITH BEST AVAILABLE CONTROL TECHNOLOGY

     Achievable emission levels for SOX and hydrocarbons with best avail-
able control technology are summarized in Table III-4 for a 50,000 BPSD
FCC unit.  The achievable level is 300 ppm for FGD based on 90% removal
of SOX from flue gas with a 3,000 ppm inlet concentration level.  The
corresponding level for hydrotreating the feed is 400-500 ppm based on
desulfurization to 0.3 wt. % sulfur.
                                 -56-

-------
                                 TABLE II1-4
            ACHIEVABLE EMISSION LEVELS WITH BEST CONTROL TECHNIQUES
FCC Plant Size - 50,000 BPSD
Significant Control
Emission Point Techniques
Regenerator FGD
Feed Hydrotreating
CO Boiler
HTR
Achievable
Levels
—x
300
400-500
900 ! 2
900
(Total Feed)
Emission
- PPM
Hydrocarbon
500 1
500 !
10
10

Achievable
Rates -
SO
400
670
1.2001 2
1,000

Emission
Ib/hr
Hydrocarbon
1671
1671
3.4
2.83
      as emission without control—represents average SO  concentration
 and hydrocarbon emissions with no control.

2Based on projected national average for uncontrolled sources.

3Lower coke on catalyst with HTR produces less flue gas.

-------
          IV.  STATE AND LOCAL AIR POLLUTION CONTROL REGULATIONS
     As part of this study, we reviewed state and local air pollution regu-
lations applicable to FCC regenerators.  For reasons explained in Chapter
III, the review of emission standards was limited to SOX and hydrocarbons.
The purpose of this chapter is to determine how the emission levels required
by state and local regulations compare to the emission levels achievable
with the control systems discussed in Chapter III.

A.  SELECTION OF STATES FOR REVIEW OF REGULATIONS

    In selecting states for consideration it was important to include a
representative range of states.  We reviewed the regulations for those
seventeen states in which at least 1% (about 57,000 BPSD) of total U.S.
FCC capacity was located.  States falling into this category are shown in
Table IV-1.  As can be seen from Table IV-1, using this criteria we were
examining state regulations governing 5,297,644 BPSD, or 93% of total U.S.
capacity (fresh feed and recycle).

B.  REGULATIONS

    Table IV-2 shows the state regulations applicable to FCC's in the
seventeen states examined.

    Some of the states listed in Table IV-2 have SOX emission standards in
terms of ambient air quality or stack height.  The application of these
standards is naturally site-specific.  To establish the existing SOX emis-
sion baseline, an average emission standard was assigned to these states.
This was accomplished by completing the weighted average of the SOX emission
standards applicable to states with source emission standards as shown in
Table IV-3.  It was assumed that this weighted average SOX standard was
representative of states having standards in terms of air quality or stack
height and states not included in Table IV-2.

    The SOX emissions in the flue gas from FCC regenerators depend on feed-
stock, percent sulfur in the feedstock, and process variables.  The affect
of these variables are discussed in Chapter II.  The typical SOX concentra-
tion in the regenerator flue gas (Table II-2) is 800-900 ppm.  However,
the emissions may be as high as 1,500-2,500 ppm.  A few states regulate
SOX emissions to 500 ppm or less.  The remaining states have no SOX emis-
sion regulation, or control it to 2,000 ppm.  Flue gas desulfurization
(FGD) systems and hydrotreating units have economic sulfur removal efficiency
up to 90%.  The SOX emissions from FGD systems range from 50-300 ppm (de-
pending on inlet SOX concentration).  The SOX emissions from FCC regenerators
using hydrotreated oil (0.3% S) range from 400-500 ppm.  Thus, the states
having stringent regulations may force the FCC regenerator owners to use
low sulfur feedstocks (blended or hydrotreated) or to desulfurize the flue
gas.  Obviously, states having an SOX emission standard equal to 2,000 ppm
have less control.  The reduction in emissions from state regulations levels
achievable with existing control technology are shown in Table IV-4 for a
typical FCC unit.
                                    -58-

-------
                              TABLE IV-1




               STATES WITH FCC CAPACITY OF 1% OR GREATER
PAD
District State

1 Delaware
New Jersey
Pennsylvania
2 Illinois
Indiana
Kansas
Minnesota
Ohio
Oklahoma
3 Louisiana
Mississippi
Texas
4 Montana
Utah
Wyoming
5 California
Washington
No. of
Refineries
With FCC's

1
4
4
9
4
10
3
6
10
9
2
30
4
6
7
13
3
Capacity
(Fresh Feed
and Recycle)
(BPSD)
77,000
269,444
224,300
523,277
203,800
206,450
74,500
248,500
231,675
677,728
76,850
1,527,571
72,500
70,160
74,078
621,211
118,600
     Total
125
5,297,644
Source:   The Oil and Gas Journal,  March 29,  1976.
                                 -59-

-------
                                TABLE IV-2

                          STATE AIR REGULATIONS
PAD
District
1



State
Delaware
New Jersey
Pennsylvania

sox
Ambient1
2,000 ppm2a
500 ppm
            Illinois
            Indiana
           Kansas
            (Kansas City)
            (Wichita)
           Minnesota
           Ohio


           Oklahoma

           Louisiana
           Mississippi


           Texas
           Montana
           Utah
           Wyoming

           California
            (LA County)
           Washington
2,000 ppm
     19.5P0'673.*
(County regulation)
None

None
Existing - 2,000 ppm
New - 500 ppm

Ambient1

Existing - 2,000 ppm
Existing - 2,000 ppm
New - 500 ppm

440 ppm
None
Ambient1
Ambientl

(County regulation)
2,000 ppm25
1,000 ppm
     Hydrocarbons

None
Ambient1
None

100 ppm equivalent
methane (molecular
weight 16.0)
Waste gas stream to be
burned in direct-flame
afterburner or boiler


None

None
None


None

None
None

Waste gas stream burned
at 1,300°F or greater in
a direct-flame incinerator
or boiler

None
None
None
85Z reduction5
Ambient1
1 Ambient air quality  standard  only.

 28 Also regulation on SOX  from fuel burned  310 ppm by volume adjusted  to
  12Z CO 2 by volume  (non-commercial fuel) .
2^Also regulation on sulfur  content of  fuel 0.5% sulfur by weight.

3 Ep » Emissions  in pounds of S02/hour.
  P  » Tons of process weight/hour.
 "* Indiana also regulates  maximum hourly ground  level concentrations in
  various air pollution regions of the  state.   The regulation  is expressed
  as a formula involving:

  Sp « Pounds of  sulfur emitted/ ton of  process  weight  input.
  n  - Number of  stacks.
  a  - Plume rise factor  of  0.7.
  hs - Stack height in feet.
  P  - Total process equipment  capacity weight  input,  tons/hour.

  C    "       °'75    °-25
                ahs

5 Hydrocarbon emissions  not  to  exceed  15 Ib/day or  3  Ib/hour.
                                     -60-

-------
                               TABLE IV-3
                    STATE AIR POLLUTION STANDARD FOR
                    0  EMISSIONS FROM FCC REGENERATOR
    State

New Jersey

Pennsylvania

Illinois

Indiana

Ohio

Louisiana

Mississippi

Texas

Washington

California

     Total

Weight Average
FCC Capacity
 % of Total
  Capacity
   79.14
Emission Standard
 Existing Source
 ppm in Flue Gas
Emission Standard
   New Source
 ppm in Flue Gas
4.75
3.95
9.22
3.59
4.38
11.94
1.35
26.92
2.09
10.95
2,000
500
2,000
7201
2,000
2,000
2,000
440
1,000
2,000
2,000
500
2,000
7201
500
2,000
500
440
1,000
2.000
                       1,337
                            1,227
1Based on average size of FCC capacity in the State of Indiana.
                                   -61-

-------
    Two states, Indiana and Texas, require combustion of flue gas at a
certain temperature (usually 1,300°F) to control hydrocarbon emissions.
Residence times on the order of 0.3 sec are also required.  The hydrocarbon
emissions from combustion of flue gas are usually lower than 10 ppm.  Two
other states regulate hydrocarbon emissions from FCC regenerators (see
Table IV-5).  States with regulations contain 40% of the FCC capacity.  No
hydrocarbon emission regulations were assumed in the remaining states.

    The FCC units with CO boilers have very low hydrocarbon emissions.
This is apparently also true for units employing HTR technology, although
stack test results are scanty (Appendix A).  The best practicable technology
to control hydrocarbon emissions is to employ HTR or CO boilers as mentioned
in Chapter III.  It is estimated that only 28% of the FCC capacity will
not be equipped with either of the above two options in 1975 (Chapter V).
The effective emission level under state regulations (only a few states
have hydrocarbon regulations) and best practicable technology are the same
in this case.
                                    -62-

-------
                                                    TABLE IV-4
                                      SOV AND HDYROCARBON EMISSION REDUCTION
Emission Source    Control Technique
Regenerator
FGD

Feed Hydrotreating



CO Boiler


HTR
FROM TYPICAL PLANT BASED ON ACHIEVABLE CONTROL
FCC Plant Size - 50,000 BPSD
Emission Rate
w/ Achievable Control
lue Ib/hr
SO * Hydrocarbon2
400 167
zing 6702 167
1,200 3.4
1,000 2.8
(Total Feed)
Emission Rate
w/ State-Required
Control (Average)
Ib/hr
SO Hydrocarbon
1,200 167
1,200 167
1,200 167
1,000 167

Emission Reduction
By Applying
Achievable Control
Ib/day
SO Hydrocarbon
19,200 0
12,720 0
0 3.926
0 3,941
           expressed as S0_.
                          «
      2Based on 500 ppm SO. in regenerator flue gas consistent with  'economic1  removal  to  0.3%  S  in FCC  feed.

-------
                              TABLE IV-5

              STATE REGULATIONS FOR HYDROCARBON EMISSIONS


                          Capacity
                        (Fresh Feed
  State                 and Recycle)                   Control Level
                          (BPSD)                            (ppm)

Illinois                   523,277                          100

Indiana                    203,800                 10 (from Incinerator)

Texas                    1,527,571                 10 (from incinerator)

Wisconsin                   10.700                      85% removal

     Total               2,265,348*
*2,265,348 BPSD represents 40% of total U.S. FCC capacity,
                                   -64-

-------
                     V.  ESTIMATED EMISSION REDUCTIONS
     Several models have been developed by the EPA for the determination of
regulatory priorities over the past few years.  The TRC Model IV, which is
applied here, provides a quantitative estimate of the anticipated impact of
standards of performance in preventing atmospheric emissions.

     The additional control potential of new or revised standards of per-
formance stems from the application of emission standards that are more
stringent than those presently applied to construction and modification.
This potential for a specified time period is expressed as (Tg - TJJ) where

     Ts * emissions under 1975 state control regulations (tons/yr)

     TN - emissions under new or revised standards of performance (tons/yr)


     To calculate the control potential of standards of performance, other
factors must be considered such as total growth in capacity, the portion
of growth requirements that can be satisfied from present unused capacity,
and the obsolescence and replacement rates of existing facilities.  TRC
Model IV is a mathematical relationship incorporating these parameters,
which can be used to determine the quantity (Ts -
     A detailed description of Model IV is provided in Appendix E.  Param-
eters requiring quantitative information are defined below.

A.   MODEL IV PARAMETERS

1.   Parameter Definitions
     LU
  total emissions in the impact year assuming no control (tons/yr)

• total emissions in baseline year under baseline year regulations
  (tons/yr)

• normal fractional utilization rate of existing capacity, assumed
  constant during time interval

- baseline year production capacity (production units/yr)

* construction and modification rate to increase industry capacity
  (decimal fraction of baseline capacity/yr)

- construction and modification rate to replace obsolete capacity
  (decimal fraction of baseline capacity/yr)

= production capacity from construction and modification to in-
  crease output above baseline year capacity (production units/yr)

= production capacity from construction and modification to replace
  obsolete facilities (production units/yr)
                                    -65-

-------
     Eg. « allowable emissions from existing sources under existing regula-
           tions (mass/unit capacity)

     Eg  » allowable emissions from new and modified sources under existing
       2   regulations (mass/unit capacity)

     EU  * allowable emissions from new sources under NSPS (mass/unit capa-
           city)

2.   Baseline Year

     The model is applied to quantitatively estimate the emission reduction
due to the enforcement of anticipated standards.  The pollutants of concern
are hydrocarbons and oxides of sulfur, for which there are no NSPS at pres-
ent.  The baseline year of 1975 is used to determine the effect of best
control technology.  The model is used to estimate the effect of antici-
pated NSPS over the period of ten years from 1975 to 1985.  Therefore,
1985 is the impact year.

     The numerical values assigned to the parameters are discussed below.

     K  - A value of 0.97 was used based on typical operating experience
          (see discussion in Chapter 1).

     A  - The baseline FCC production capacity of 5,670 KBPSD (2069.6 MBPY)*
          is based on the latest Oil and Gas Journal refinery survey and
          represents the existing capacity at the end of 1975.

     PQ - The compound rate of capacity growth through 1980 is 0.015 (deci-
          mal fraction of 1975 baseline capacity).  Thereafter, PC is
          assumed to be negligible (0.0).  The basis for these projections
          is discussed in Chapter I.

     PB - The simple rate or replacement of obsolete capacity used in the
          calculation is 0.025 (decimal fraction of 1975 baseline capacity).
          This is based on an estimated 40-year equipment life which is
          more than twice the IRS depreciation life of 16 years.  However,
          there are FCC units still operating that were built during World
          War 11.  Considering present uncertainties regarding energy
          policy and oil company devestiture, most refiners will continue
          to repair existing units in the short-term future.

     B&C - The FCC capacity ultimately affected by NSPS was established by
          using the above factors to construct Figure V-l.  The values of
          B and C at the end of the ten-year study period are 1,420 KBPSD
          (356.2 MBPY) and 440 KBPSD (160.6 MBPY), respectively.  There
          are no FCC capacity additions assumed after 1980.  The replace-
          ment of obsolete capacity is based on a 2.5% simple growth rate
          using the 1975 capacity as a basis.
*K » thousand
 M - million
                                   -66-

-------
                             Replacement  rate based  on 40 year life
                             < I I t * I '  I  ' '  • I i ; t -1 I t i ' i i • t i i  i i  -i  • i i i
                             Source:  ADL  Estimates
                  19 8Q

                  J-M'
                    Years
 FIGURE V-l   PROJECTED FCC
CAPACITY ADDITIONS AND REPLACEMENTS

  AFFECTED BY NSPS
!!n  Mil!
                             -67-

-------
     Values for the following parameters are specific to the pollutant
being assessed.

     Eg - The uncontrolled SOX emission factors were obtained from the
          knowledge of FCC capacity in each state, feed hydrotreating
          capacity in each state, sulfur content of feedstocks, the mix
          of oil feedstocks used in each PAD district, and relationship
          between SOX in the regenerator flue gas and sulfur content in
          the feed.  The details of this calculation are given in section
          B of this chapter.  The EJJ obtained for SOX are 502 and 551
          lb/1,000 barrels in 1975 and 1985, respectively.  The correspond-
          ing values of EU for hydrocarbons are 23.0 and 11.5 lb/1,000
          barrels.  The decline in emissions is due to increased use of
          HTR between 1975 and 1985.

     ES - The emission factors under existing regulations were obtained
          by applying state regulations to emissions obtained from the
          uncontrolled sources in each state.  The details of this calcu-
          lation are also given in section B.  The £5 obtained for SOX
          is 401 lb/1,000 barrels in 1975.  The values of Es. and ES~ are
          445 and 439 lb/1,000 barrels in 1985.  The Eg values for hydro-
          carbons are the same as Eg since few states have standards and
          many units already have CO boilers.

     EN - The SOX emission factors under best control technology were ob-
          tained for two cases:  1) outlet SOX concentration limited to
          300 ppm, and 2) outlet SOX concentration limited to 500 ppm.
          Necessary correction factors were applied to emission factors
          in some cases where the existing state emission regulations were
                                                    EM e<
                                                   Lb/l,(
             for hydrocarbons is 1.56 lb/1,000 barrels (as methane) based
          on typical emissions from CO boilers and HTR.
lower than the above concentrations.  The E» equivalent to the
above SOX control levels are 182 and 290 lb/1,000 barrels.  The
B.   APPLICATION OF THE MODEL

     The model is applied to determine emission reduction of SOX and hydro-
carbon pollutants.  Since some parameters defined earlier are different
for each pollutant, the application of the model to SOX emissions and to
hydrocarbon emissions are discussed separately.

1.   Application of TRC Model IV to SO  Emissions

     In applying the model to determine the impact of NSFS for SOX we
accounted for various factors that affect the sulfur emission from FCC
regenerators.  These include:

     1.  variations in FCC feed sulfur content by geographic
         region and time,
     2.  incidence of feed hydrodesulfurization,
     3.  recycle ratio, and
     4.  existing state emission regulations.
                                    -68-

-------
     The use of these factors to develop Ey, £5, and Eg are discussed next
followed by the calculation of total emissions of SOX and TJJ - Tg.

     a.  SO  Concentration in Flue Gas from FCCU Regenerators

     The regenerator outlet SOX concentrations shown in Table V-l are based
on the correlation presented in Figure V-2.20  In Figure V-2, the outlet
SOX concentration is related to sulfur content of the feedstock for dif-
ferent types of feed.  This relationship is based on 2% excess oxygen in
the flue gas.  The SOX emissions in the flue gas from regenerators for the
case of hydrotreated feed were also obtained from Figure V-2 and are con-
sistent with other literature.*'22  The average sulfur content of hydro-
treated feeds was assumed at 0.3%, to be consistent with economic desul-
furization levels.

     In each PAD district, the blend of gas oil feeds to the FCC differs.
This is shown in Table V-2.  The mix also varies with time, and the break-
down by source is given for the years 1975 and 1985.  The mix does not In-
clude hydrotreated oil; this is dealt with separately in the preceding
sections.  The average SOX concentrations In the regenerator flue gas for
each PAD district is obtained by weighting the emission data in Table V-l.

     In FCC operation, part of the light cycle oil is recycled.  An overage
recycle rate equal to 10% of the total FCC feed was assumed.  In Table V-3,
SOX emissions from the recycle oil are developed based upon 100Z recycle
oil.  The average weight percent sulfur in FCC fresh feed was obtained
from the type of mix (Table V-2) and the sulfur content in different types
of gas oils (Table V-l).  The weight percent sulfur in the recycle oil was
obtained from a published correlation of sulfur content in the light cycle
oil and sulfur content of FCC feed.20  The SOX concentrations in the flue
gas from the recycle oil was then estimated using Figure V-3.

     The estimated average SOX concentrations in the flue gas from FCC
regenerators for each state are given in Table V-4 with appropriate weight-
ing for fresh, recycle and hydrotreated feeds.  The SOX concentrations are
given for the years 1975 and 1985.  The distribution of FCC and hydrotreat-
ing capacity in each state was assumed to be constant over the period 1975
to 1985.  The 1975 FCC and the hydrotreating capacities were obtained from
a recent survey.9  An SOX concentration equal to the average SOX concentra-
tion was used to fill the data gap (PAD District IV, 4.2% FCC capacity).

     b.  SO  Emissions with No Control. E^ SO

     The average uncontrolled SOX concentrations from FCC regenerators
including the contributions from recycle and desulfurized feeds are given
in Tables V-5 and V-6.  The percent capacity in each state and the SOX
concentrations in the effluent were obtained from Table V-4.

     It is necessary to have a relationship between flue gas flow rate and
oil feed to FCC units to convert SOX concentrations to emission factors.
The coke yield varies with the severity of cracking.  A coke yield equal
                                    -69-

-------
                    TABLE V-l

SO  CONCENTRATION IN FLUE GAS FROM FCC REGENERATORS
A
Crude
Code1
L
T
N
A
V
0
MC
AL
CW
CV
I
AMS
HDS
HDS Recycle
FOR DIFFERENT CRUDES

S in FCC Feed (%)
0.237
1.659
0.269
1.925
1.374
0.202
0.613
0.17
1.22
1.434
0.078
0.885
0.3
0.5

SOX in Outlet
Flue Gas. ppm2
300
1,300
400
2,100
1,250
300
500
280
450
480
175
1,000
450
800
      Chapter I for explanation of code.  HDS refers
  to hydrodesulfurization oil.

 20utlet flue gas 0. concentration - 2%.
                     -70-

-------
SO. in CO boittr stock gas, pern by whim*, dry basis
i x t s § i i 1 I i

Itgend
O W Itw gas til
6 torn* (twill
Q time* CO. MOkfcM
O S« In fat «l
O d*-n*i Ml
X (gfif gm «l
® leliwty ilock iMawnmcnt
Dofkfiw^ (Minlt »rf IM l» ttmf




—;





j£r<'
\& now


S




.

X
/

v& '
$
limit
•qvtf

tor
m»n






1








>^
jf

*





4
^

^*r^

X Board
••iiflitg





^
1


lir Cor
limit
**J*V*




*


rtrol
for
••ftf
1
1.1 U »J 1.4 ».5 1.0 tl M
Stfffuf CMtMl OT FCC MM, Wt %
   FIGURE V-2   SO  CONTENT OF CO BOILER STACK GAS*
^Assuming no supplementary fuel to the boiler and 2.0Z
 oxygen in boiler stack gas.


Source:  Gulf R&D.20
                          -71-

-------
                               TABLE V-2
          SO  CONCENTRATION IN FLUE GAS FROM FCC REGENERATOR
PAD
District1
I




I la




lib



Ilia
Illb




V




X
Type of
Crude1
L
N
T
A
V
AL
L
T
0
MC
A
AL
L
T
0
A
MC
L
T
L
T
N
A
V
CW
CV
A
MC
I
ANS
Crude Average SOX Concentration
Volume Percent1 in Outlet, ppn2
1975
^•Hfl^B^M
15.4
16.2
7.6
7.6
31.7
21.5
7.6
13.1
61.5
17.8
—
--
6.0
70.3
4.9
8.5
10.3
88.7
11.3
47.4
41.4
3.8
5.3
2.1
37.4
13.8
31.3
7.1
10.4
1985 1975 1985

19.2
••••
45.7
16.5
18.6
__
10.0
55.8
17.1
17.1
——
55.1
4.4
40.5
—


825 1,300




465 700




1,175 1,575

88 • 2 } ... ,2-
11.8 / *15 *"
47.4
41.4
3.8
5.3
2.1
40.0
13.2
~—
—


825 825




945 700

46.8
     Chapter I for details.
2The SO  concentrations in the flue gas are based on no hydrotreating of
 feed to catalytic reactor.  The SOX concentrations are based on 2% 02
 concentration in flue gas from FCC regenerator system.
                                  -72-

-------
                                    TABLE V-3
                          SO  EMISSIONS FROM RECYCLE OIL
A
(100% Basis)
PAD
District*

I
Ha
lib
Ilia
II Ib
V
Wt. % S in
FCC Feed
1975
0.82
0.47
1.42
0.40
1.54
1.31
1988
1.19
0.64
1.70
0.41
1.54
1.09
Wt. % S in
Recycle Oil
1975
1.10
0.74
2.02
0.70
2.20
1.92
1988
1.75
0.90
2.40
0.70
2.20
1.50
SOX Concentration
in Emissions
1975
1,025
875
1,400
800
1,500
1,300
1988
1,250
1,000
1,650
800
1,500
1,200
*See Chapter I for details on classification by PAD Districts.
                                      -73-

-------

a
a.
a
CO
c
o
- 1
*rt
CO
co
•g
w
c
c
o
•H
u
«
M
u
C
01
o
c
0
u
ox
y^

1600
1400


1200 •



1000


800

600 •




400 •

200 -
0
                      Based on SoCal and Gulf Data20*21
                     Sulfur in Recycle Oil, Wt. %
FIGURE V-3   SO^ EMISSIONS ATTRIBUTABLE TO SULFUR
               A
IN RECYCLE OILS

-------
TABLE V-4
SO CONCENTRATION IN UNCONTROLLED EMISSIONS

PAD
District
I























II
(Large)








A

State
Delaware
Fresh
Recycle
Total
New Jersey
Fresh
Recycle
Total
New York
Hydro treat ing
Recycle
Fresh
Recycle
Total
Pennsylvania
Hydrotreating
Recycle
Fresh
Recycle
Total
Virginia
Fresh
Recycle
Total
Illinois
Hydrotreating
Recycle
Fresh
Recycle
Total
Indiana
Fresh
Recycle
Total
Capacity
BPSD
(1975)

69,300
7,700
77,000

242,500
26,944
269,444

20,000
2,200
22,300
2,500
47,000

40,000
4,400
161,870
18,030
224,300

28,800
3,200
32,000

27,000
3,000
443,950
49,327
523,277

183,420
20,380
203,800
sox

(1975)

745
105


745
105


190
40
390
55


80
15
595
85


745
105


25
5
1,000
135


1,060
140

Concentration
in Flue Gas
(1985)

1,170
125
850

1,170
125
850

190
40
615
65
675

80
15
940
• 100
775

1,170
125
850

25
5
1,335
155
1,165

1,420
165
1,200






1,295



1,295





910





1,135



1,295





1,520



1,585
   -75-

-------
               TABLE V-4 cont'd

PAD
District














II
(Small)






















State
Kentucky
Fresh
Recycle
Total
Michigan
Hydrotreating
Recycle
Fresh
Recycle
Total
Tennessee
Fresh
Recycle
Total

Kansas
Fresh
Recycle
Total
Minnesota
Hydrotreating
Recycle
Fresh
Recycle
Total
Missouri
Fresh
Recycle
Total
Nebraska
Fresh
Recycle
Total
North Dakota
Fresh
Recycle
Capacity
BPSD
(1975)

49,500
5,500
55,000

12,500
1,390
29,710
3,300
46,900

12,150
1,350
13,500


185,800
20,650
206,450

20,000
2,200
47,050
5,250
74,500

47,700
5,300
53,000

2,610
290
2,900

30,600
3,400
SO
X
(1975)

1,060
140


120
25
745
100


1,060
140



420
90


120
25
295
60


420
90


495
90


420
90
Concent rat ion
in Flue Gas
(1985)

1,420
165
1,200

120
25
995
115
990

1,420
165
1,200


630
100
510

120
25
440
70
500

630
100
510

630
100
510

630
100
                                                           1,585
                                                           1,255
                                                           1,585
                                                             730
                                                             655
                                                             730
                                                             730
Total
34,000
                                           510
730
                       -76-

-------
TABLE V-4 cont'd

PAD
District














Ilia















Illb












State
Oklahoma
Fresh
Recycle
Total
Wisconsin
Fresh
Recycle
Total
Ohio
Hydrotreating
Recycle
Fresh
Recycle
Total
Arkansas
Fresh
Recycle
Total
Louisiana
Hydrotreating
Recycle
Fresh
Recycle
Total
Mississippi
Hydrotreating
Recycle
Fresh
Recycle
Total
New Mexico
Fresh
Recycle
Total
Texas
Hydrotreating
Recycle
Fresh
Recycle
Total

Capacity
BPSD
(1975)

208,507
23,168
231,675

9,630
1,070
10,700

22,500
2,500
201,150
22,350
248,500

16,200
1,800
18,000

41,000
4,560
568,960
63,210
677,730

23,000
2,560
46,170
5,130
76,850

15,800
1,760
17,560

70,000
7,800
1,304,810
144,960
1,527,571
-77-
sox

(1975)

420
90


420
90


40
10
375
80


375
80


30
5
350
75


135
25
250
55


740
150


20
5
705
140


Concentration
in Flue Gas
(1985)

630
100
510

630
100
510

40
10
565
90
505

385
80
455

30
5
355
75
460

135
25
255
55
465

745
150
890

20
5
705
140
870

                                         730
                                         730
                                         705
                                         465
                                         465
                                         470
                                         895
                                         870

-------
                      TABLE V-4  cont'd

PAD
District
IV













State
Colorado
Fresh
Recycle
Total
Montana
Fresh
Recycle
Total
Utah
Fresh
Recycle
Total
Capacity
BPSD
(1975)

21,500
2,400
23,900

65,250
7,250
72,500

63,144
7,016
70,160
SOX

(1975)

725
80


725
80


725
80

Concentration
in Flue Gas
(1985)

820
90
805

820
90
805

820
90
805
                                                              910
                                                              910
                                                              910
Wyoming
  Hydrotreating
  Recycle
  Fresh
  Recycle
  Total

California
  Hydrotreating
  Recycle
  Fresh
  Recycle
  Total
16,000
1,780
50,670
5,630
190
20
550
60
 74,078
58,000
6,400
501,090
55,721
40
10
760
120
621,211
Hawaii
Fresh
Recycle
Total

20,700
2,300
23,000

850
130

Washington
  Hydrotreating     8,500
  Recycle             940
  Fresh            98,240
  Recycle          10,920
  Total
            35
             5
           785
           120
805
930
                                             980
                            195
                             20
                            625
                             70
                             40
                             10
                            565
                            105
                                                       630
                                                       120
           35
            5
          580
          110
910
720
                                   750
118,600
945
730
                            -78-

-------
                                  TABLE V-5
                  ESTIMATED SO  EMISSIONS FROM FCC IN 1978
Uncontrolled
PAD
District State
I Delaware
New Jersey
New York
Pennsylvania
Virginia
II (Large) Illinois
Indiana
Kentucky
II (Small) Kansas
Michigan
Minnesota
Missouri
Nebraska
North Dakota
Ohio
Oklahoma
Tennessee
Wisconsin
Arkansas
III Louisiana
Mississippi
New Mexico
Texas
IV Colorado
Montana
Utah
Wyoming
V California
Hawaii
Washington

Percent
Capacity
1.36
4.75
0.83
3.95
0.32
9.22
3.59
0.97
3.64
0.83
1.31
0.93
0.05
0.60
4.38
4.08
0.24
0.19
0.32
11.94
1.35
0.31
26.92
0.42
1.28
1.24
1.31
10.95
0.41
2.09

Emissions
SO Emissions Existing Regulations
PPrc
850
850
675
775
850
1,165
1,200
1,200
510
990
500
510
510
510
505
510
1,200
510
455
460
465
890
870
805
805
805
805
930
980
945

tons/yr
7,499
26,191
3,634
19,858
1,765
69,679
27,946
7,551
12,043
5,330
4,249
3,077
165
1,985
14,349
13,498
1,868
629
945
35,629
4,072
1,790
151,928
2,193
6,684
6,475
6,841
66,060
2,607
12,812
519,352
tons/yr
7,499
26,191
3,634
12,812
1,765
69,679
16,768
\7',923l
585
5,330
4,249
3,077
165
1,985
(14,349
U4.5801
13,498
1,868
629
945
35,629
4,072
1,790
76,837
2,193
6,684
6,475
6,841
66,060
2,607
(12,812
117, 0731
(414, 5792
\463,7693
1Emissions under state new source regulations assuming all capacity affected.
2Total emissions from all sources with existing source regulations.
3Total emissions from all sources with new source regulations.  Note that for
 some states the new and existing source regulations are identical.
The SOX emissions are reported as S02-
                                    -79-

-------
                                   TABLE  V-6
ESTIMATED SO


A
EMISSIONS

Uncontrolled

State
Delaware
New Jersey
New York
Pennsylvania
Virginia
Illinois
Indiana
Kansas
Kentucky
Michigan
Minnesota
Missouri
Nebraska
North Dakota
Ohio
Oklahoma
Tennessee
Wisconsin
Arkansas
Louisiana
Mississippi
New Mexico
Texas
Colorado
Montana
Utah
Wyoming
California
Hawaii
Washington


Percent
Capacity
1.36
4.75
0.83
3.95
0.32
9.22
3.59
3.64
0.97
0.83
1.31
0.93
0.05
0.60
4.38
4.08
0.24
0.19
0.32
11.94
1.35
0.31
26.92
0.42
1.28
1.24
1.31
10.95
0.41
2.09


SOV Emissions
PPm
1,295
1,295
910
1,135
1,295
1,520
1,585
730
1,585
1,255
655
730
730
730
705
730
1,585
730
465
465
470
895
870
910
910
910
910
720
750
730


tons/yr
11,940
41,703
5,121
30,395
2,809
95,012
38,577
18,015
10,423
7,062
5,817
4,603
247
2,969
20,935
20,192
2,579
940
1,009
37,641
4,302
1,881
158,781
2,591
7,897
7,650
8,082
53,450
2,085
10,344

615,052
FROM FCC IN 1985
Emissions with
Existing State
Regulat ions
tons/yr
(11,940
\11.3131
41,703
5,121
13,390
(2,809
(2.6611
95,012
17,524
18,015
(8,792
18.0691
(7,062
16.9041
5,817
4,603
247
2,969
(20,935
114, 8481
20,192
2,579
940
1,009
37,641
4,302
1,881
80,303
2,591
7,897
7,650
8,082
53,450
2,085
10,344
496, 885^
489, 1423

Emissions

if S0_
in All Flue Gas is**
<300 ppm
tons/yr
2,766
9,661
1,688
8,034
651
18,752
7,302
7,403
1,973
1,688
2,664
1,892
102
1,220
8,909
8,298
488
386
651
24,285
2,746
631
54,752
854
2,609
2,522
2,664
22,271
834
4,251
202,947

<500 j>pm
tons/yr
4,610
16,102
2,814
13,390
1,085
31,254
12,169
12,339
3,288
2,814
4,440
3,153
169
2,034
14,848
13,830
814
644
1,009
37,641
4,302
1,051
80,303
1,424
4,339
4,203
4,441
37,119
1,390
7,085
324,104

1Emissions under state new source regulations assuming all capacity affected.
2Total emissions from all sources with existing source regulations.
3Total emissions from all sources with new source regulations.   Note that for
 some states the new and existing source regulations are identical.  The SOX
 emissions are reported as 862.
''Based on all capacity affected  by NSPS.  Used to calculate weighted average
 Ey,  not ultimate impact.

                                     -80-

-------
to 6 percent and 5 percent is assumed for conventional regenerator and high
temperature regenerator, respectively.  The capacity of a FCC having HTR
is assumed to be 15% for 1975 and 60% for 1985.

     The air to coke mass ratio equal to 14.2 is used to determine the
amount of flue gas with 2 percent excess oxygen.  A flue gas to oil mass
equal to 0.936 and 0.782 is obtained from the above bases for conventional
regenerators and HTR, respectively.  The average flue gas to oil ratio cor-
responds to 0.91 and 0.84 for the years 1975 and 1985, respectively.

     The FCC capacity, SOX concentrations in the flue gas, and mass ratio
of flue gas to FCC feed oil are used to determine uncontrolled emissions
in Tables V-5 and V-6.  An FCC feed oil density equal to 315 Ib/barrel is
used in the above calculation.  From Tables V-5 and V-6, the uncontrolled
emissions (national average) in the years 1975 and 1985 are 519,352 ton/yr
and 615,052 ton/yr, respectively.  The emission factor calculated from
this is 501.9 lb/1,000 barrels (805 ppm) in 1975 and 551.6 lb/1,000 barrels
(905 ppm) in 1985.

     c.  SO  Emissions with Existing Regulations, E , SO

     The state emission standards summarized in Chapter IV are used as
existing standards.  The emissions under existing regulations for FCC's
are developed in Tables V-5 and V-6.  The emissions shown represent the
lesser of the uncontrolled emissions or allowed emissions.  The determi-
nation of allowable emissions was obtained from allowable SOX concentration
in the flue gas (see Chapter IV), mass ratio of flue gas to oil feed to
FCC, and the capacity of FCC.  The total estimated SOX emissions (ton/yr)
under  existing regulations are shown in Tables V-5 and V-6.  The calcu-
lated emission factors Eg (1975), ESl (1985), and Eg2 (1985) are 400.7,
445.6, and 438.7 lb/1,000 barrels, respectively.

     Here, Eg.. (1985) refers to allowable (national average) SOX emissions
under existing regulations for existing FCC's in the year 1985, Eg-
(1985) refers to allowable SOX emissions allowing for weighting of those
states which have different standards for new or modified FCC in the year
1985, and Eg (1975) refers to allowable emissions in the year 1975.

     d.  SO  Emissions with Anticipated Regulations, E^
           it                                          W

     The EJJ  is  determined for two potential control levels:  (1) 300 ppm
SOX and (2) 500 ppm SOX in the flue gas.  It may not be possible to use
low sulfur oils without a FGD system to meet the 300 ppm standard.  The
500 ppm level can very likely be achieved by using low sulfur oil (maybe
hydrotreated) as FCC feed.

     Since some states have emission regulations that fall below the 500
ppm level, a weighted average EJJ was determined by calculating the total
SOX emission using the lesser quantity for each state and then dividing
the total FCC capacity.  The emission factors calculated from the total
emissions given in Table V-6 are E^ 30Q = 182 lb/1,000 barrels oil and
                                    -81-

-------
EN 500 = 290.7 lb/1,000 barrels oil.  It should be noted that EN 500 corre-

sponds to less than 500 ppm SOX In the flue gas.  This is because the SOX

regulations for some states are lower than 500 ppm.



     e.  Calculation of Total SOX Emissions

         and Expected Reduction Due to NSPS



     •  Total SO  emissions in 1975, T
                X                     **


        T  - E- • K • A
         A    S




             " i°000 x °'97 x 5'67 x 1()6 * 2»203»809 Ib/day




             - 402,195 ton/yr



     •  Total SO  emissions in 1985, assuming no control, T
               »n * 0.97 x (5.67 x 106 + 0.44 x 106)
             J.,UUU



           - 3,269,170 Ib/day



           - 596,620 ton/yr



     •  Total SO  emissions in 1985 under existing standards, T
                X                                              o


        T  - E  K(A-B) + E  (B+C)
         S    S1          S2





           - I7§5§ x °'97(5'67 x lo6 - x-42 x lo6) + 17555 x


             0.97(1.42 x 106 + 0.44 x 106)



           - 1,836,986 + 791,502 Ib/day



           « 479,699 ton/yr



     •  Total SO  emissions in 1985 under anticipated NSPS, T
                X                                            "


        Two levels of NSPS are assumed:  (1) 300 ppm and  (2) 500  ppm.




        TN,300 ' £
                 1,836,986 + rr x 0.97(1.42 x  106 + 0.44  x  106)
                 2,165,350  Ib/day



                 395,176  ton/yr
                                    -82-

-------
     Similarly, TM ,nn = 2,361,466 Ib/day
                 N, _>UU
                TN,soo = 430»967

     •  Emission Reduction Potential from 1975 to 1985


        TN,300 ~ TS = 84'523 ton/yr


        TN,500 - TS ' 48'732 ton/yr

2.   Application of TRC Model IV to Hydrocarbon Emission

     In this section, the estimated reduction of hydrocarbon pollutants due
to anticipated NSPS is calculated.  Model IV is applied to hdyrocarbons
in the same way as it was applied to oxides of sulfur.  The values of K,
A, B, C, PD and P. are the same in both cases.  A 1975 baseline year is
   ,      o      t.
used.

     a.  Hdyrocarbon Emissions with No Control, EL., Hydrocarbon

     Hydrocarbon emissions from the conventional regenerators are <500 ppm.
This is equivalent to 3,900 Ib/day (calculated as methane) from a 50,000
BPSD FCC unit.  Thus, the hydrocarbon emission factor for a conventional
regenerator is 78 lb/1,000 barrel.

     In 1975, 15% of the total FCC capacity was estimated to have HTR and
67% of the total FCC's capacity was estimated to have CO boilers.  The
hydrocarbon emissions in the flue gas from HTR or CO boilers is <10 ppm.
The percent of FCC having both CO boilers and HTR is not available.  By
random distribution, the percentage of FCC's having CO boilers and HTR
units is estimated as follows:

                     Percent = 100(0.15)(.67) = 10

Therefore, the percent of FCC's having CO boilers or HTR is 15 + 67 - 10 = 78.
The emission of hydrocarbon from uncontrolled units is 78 Ib/day (calcu-
lated as methane) from a 50,000 BPSD FCC unit.  Thus, the emission factor
from HTR or CO boilers is l/50th of this or 1.56 lb/1,000 barrels.

     The average hydrocarbon emission factor for all FCC is 0.28(78) +
0.72(1.56) or 22.96 lb/1,000 barrels in 1975.

     It is expected that the percent of HTR units will increase to 60 in
1985.  Therefore, by the above procedure, 87 percent of the FCC's will be
equipped with HTR or CO boilers.  Thus, the hydrocarbon emission factor,
E, is expected to decrease to 11.5 lb/1,000 barrels.
                                    -83-

-------
     b.  Allowable Hydrocarbon Emissions
         Under Existing Regulations. ECT

     The hydrocarbon emissions from FCC's are effectively controlled by
HTR or CO boilers.  HTR or CO boilers on the FCC's to increase the yield
or to improve heat recovery.  As shown in Table IV-4, 40 percent of FCC's
are located in the states where hydrocarbon emissions are controlled.  How-
ever, it is difficult to segregate FCC's which have installed HTR or CO
boilers to meet pollution regulations or other reasons.  However, the EU
was calculated based on an estimate of the national average population of
CO boilers and FCC's with HTR.  Thus,

              Eg - EJJ - 22.96 lb/1,000 barrels in 1975, and


              Eg - EU - 11.5 lb/1,000 barrels in 1985.


     c.  Allowable Hydrocarbon Emissions under
         Standards of Performance, E.. _

     The best available control technology, BACT, for hydrocarbon emission
reduction is the installation of HTR or a CO boiler in the FCC.  Thus, the
hydrocarbon concentration in the flue gas is <10 ppm; the hydrocarbon emis-
sion factor is 1.56 lb/1,000 barrels (calculated as methane).

     d.  Calculation of Total Hydrocarbon Emissions and (T  - T )

     •  Total hydrocarbon emissions in 1975, T>
        TA
                   x 0.97 x 5.67 x 106 = 126,278 Ib/day
           « 23,046 ton/yr

        Total hydrocarbon emissions in 1985 assuming no control,
              11>5 x 0.97(5.67 x 106 + 0.44 x 106)
             1,000

           * 68,157 Ib/day

           - 12,439 ton/yr

     •  Total hydrocarbon emissions in 1985 under existing standards, Tg

        Tg = Ty = 12,439 ton/yr
                                    -84-

-------
        Total hydrocarbon emissions In 1988 under anticipated NSPS, T».

        It Is expected that all the new or modified FCC's will be equipped
        with HTR or CO boilers.  Thus, TN - Tg - 12,439 ton/yr.

     •  Emission Reduction

        TS - TN - 0

        It is expected that the NSFS for hydrocarbon emissions from FCC
        regenerators will have minimal effect.

C.   IMPACT OF SULFUR AND HYDROCARBON NSPS FOR FCC REGENERATOR

     The impact of NSPS for sulfur oxides and hydrocarbon emissions from
FCC regenerators is summarized in Table V-7.  The reduction in sulfur emis-
sions by 1985 are 49,000 or 85,000 ton/yr depending on whether the control
level is 500 or 300 vppm SOX in the treated gas.  The magnitude of the
impact is small due to the relatively small growth rate of new capacity and
the low rate of replacement of existing facilities.  Another factor affect-
Ing the magnitude is that U.S. refineries run blended feedstocks from var-
ious crudes, which tends to reduce the-average sulfur in the FCC feed on
the national basis.

     The impact of NSPS for hydrocarbons is expected to be negligible since
it is unlikely that a new FCC unit would be built without either a CO boiler
or the capability for HTR.  Therefore, the practical NSPS level would be
consistent with what would occur in any event.  Stated mathematically, E^
is equal to £§« where Ego is the effective value not the statutory value,
since by that time most new units would either have CO boilers or have been
installed with HTR.
                                   -85-

-------
                          TABLE V-7
SUMMARY OF INPUT/OUTPUT VARIABLES
FOR MODEL IV
Common Factors






Changing
"o
Eu
Es
ESI
Es2
EJJ
Es2
TA
Ts
TN






Factors
(1975)
(1985)
(1975)
(1985)
(1985)

(1985)
(1975)
(1985)
(1985)
T -T
S N
1 1
2 SO
Barrel »
Outlet
K
A (1975)
PB
B (1985)
pc
C (1985)
Units
lb./l,000 Barrel
lb/1,000 Barrel
lb/1,000 Barrel
lb/1,000 Barrel
lb/1,000 Barrel
lb/1,000 Barrel
lb/1,000 Barrel
1,000 Ton/Yr
1,000 Ton/Yr
1,000 Ton/Yr
1,000 Ton/Yr
315 Ib
ppm 300
0.97
5.67 x
1.5% up
0.44 x
2.5% of
1.42 x
SO
502
551
401
445
439
1822 or
1822 or
402
480
3952 or
85 2 or



106 BPSD1
to 1980 and 0
106 BPSD
1975 capacity
106 BPSD


after



Hydrocarbon
22
11
22
11
11
2903 1
2903 1
23
12
4313 12
493 0


.96
.5
.96
.5
.5
.56
.56






3 SO  Outlet ppm 500
                              -86-

-------
                    VI.  MODIFICATION & RECONSTRUCTION
     Fluid catalytic cracking technology as it exists today is a result of
a series of evolutionary process improvements which have increased the yields
and capacity of existing units.  The significant developments which have been
responsible for these improvements include:

     •  Introduction of zeolite catalysts and riser cracking, and

     •  Raising operating pressure.

In general, most older units have been modified to take advantage of the
expansion opportunities available from the operating improvements listed
above.  Immediate projects do not indicate that another round of expansions
based on catalyst improvements is imminent.

     The airborne emissions from FCC units with conventional regeneration
are directly related to the yield of coke which deposits on the catalyst
and is removed by air oxidation (regeneration).  The affect of process
evolution on coke yield is clearly illustrated in Figure VI-1.  The intro-
duction of improved catalysts and regeneration techniques has reduced the
coke yield at a given conversion level.  However, it does not necessarily
follow that the emissions from regeneration were similarly reduced.  The
major advantage of these process improvements was that conversion could be
increased for a given amount of coke on the catalyst.  Therefore, in
practice the coke yields and hence the emissions would not have been greatly
reduced by past improvements (high temperature regeneration excepted).

     The most recent FCC technology development is complete combustion
of CO from catalyst regeneration, otherwise known as high temperature
regeneration (HTR) or in-situ CO combustion.  The objective is to obtain
more complete combustion of the coke on the catalyst.  Both thermal and
catalytic methods  are used to induce the oxidation reactions.  The major
benefits of the process include:

     •  Increased catalyst activity due to improved regeneration
        (reduced coke), and

     •  Reduced CO emissions.

     Because for a given set of conditions, there is less coke on the
catalyst with HTR, cracking severity or oil/catalyst ratios can be
increased.  The result in the first case is to improve the yields of
gasoline and distillate from a given amount of feed, and in the second
case to increase the fresh feed capacity.  Hence, there are economic
incentives which accrue to HTR in addition to the reduction of CO
emissions.  Hence, revamping FCC units for HTR is the current trend in
technology improvements.
                                   -87-

-------
                10
oo
00
I
                 8
 »
O
_J
UJ

>

UJ


O
O
                                                                  CONSTANT AIR

                                                             BLOWER OUTPUT
                         50
                                  60                  70

                                   CONVERSION, VOL.%
80
                           FIGURE VI-1   EVOLUTION OF FLUID CRACKING (UNIT C*, 1947-73)


            *With  constant air-blower output.

            Source:  Standard Oil Co. (Ind.)

-------
     A fundamental question regarding these modifications is whether they
constitute a major reconstruction as defined in the FR,25 since recon-
structed facilities must comply with NSPS regardless of the emission change.
Clearly, the modification standard does not apply, since the emissions from
the HTR conversion units are equal to or less than those from the pre-
conversion case.

     Part 60 of the regulations specifies that reconstruction occurs upon
replacement of components if the fixed capital cost of the new components
exceeds 50 percent of the fixed capital cost required to construct a com-
parable new facility.  By definition, "facilities" are identifiable pieces
of process equipment or individual components which, when taken together,
comprise a source.  A strict interpretation of the regulation would mean
that modifications to an FCC regenerator which cost more than 50 percent of
a new regenerator would classify as reconstructed facilities and have to
meet NSPS.  Depending on the cost, this potentially includes revamping for
HTR.

     Clearly, the cost of modifying the regenerator for HTR relative to
the cost of a new regenerator is an important relationship to evaluate.
Rough estimates of the HTR conversion costs versus new facilities costs
are shown in Table VI-1.

     The conversion cost for HTR is based upon costs associated with a
recent conversion at Exxon Company, U.S.A.  The catalyst regenerator and
air blower represents roughly 40% of the investment for a new FCC unit.
Based on this example, the cost of revamping for HTR is about 15% of the
cost of a new regenerator.  If the revamping costs were understated by a
factor of 2, it still represents only 30% of the new facility.  Hence,
FCC equipment modifications for HTR will not generally classify as
reconstructed sources, and NSPS will not apply.

     In Chapter V, it was noted that some of the projected capacity
expansions will be achieved by HTR conversions.  However, there is no
data to determine what fraction of the capacity additions will be a
result of HTR.  Since the total growth in capacity is relatively small
(<1% per year through 1985), we assumed for the NSPS impact evaluation
that all this capacity would have to meet new performance standards.
This slightly overstates the impact of NSPS, so that the results represent
a maximum case.  The alternative is to make a gross estimate of the
capacity increase due to HTR, which is an exercise in hair splitting,
since the total contribution is not very significant.
                                   -89-

-------
                          TABLE VI-1

               ESTIMATED COST OF FCC REGENERATOR
                MODIFICATIONS AND NEW EQUIPMENT

                   Capacity - 45,000 Bbl/Day


                                         New          Modified

Capital Investment1 (1975)                     $106

   New FCC Unit                         21.6

   Regenerator                           8.6            1.2

   Modified - Percent of New             —              14
1 Materials and Labor

Source;  Hydrocarbon Processing, 1974 Refining Handbook, and
         Exxon R&E.
                              -90-

-------
                                  REFERENCES


1.   Anon.,  Atmospheric Emissions from Petroleum Refineries,  A Guide for
     Measurement and Control, U.S. Department of Health, Education and Welfare,
     Public  Health Services,  Division of Air Pollution,  Publication No. 763,
     1960.

2.   Sussman,  V.H., Atmospheric Emissions from Catalytic Cracking Unit
     Regenerator Stacks, Report No. 4, Los Angeles County Air Pollution Con-
     trol District, June 1957.

3.   Danielson,  J.A., Air Pollution Engineering Manual,  Air Pollution Control
     District, County of Los  Angeles, U.S. Department of Health,  Education,
     and Welfare, Public Health Service, Bureau of Disease Prevention
     Environmental Control, National Center for Air Pollution Control,
     Cincinnati, Ohio, 1967.

4.   Shea, E.P., Source Testing - EPA Task No.  8, Standard Oil Company,
     Richmond, California, Midwest Research Institute, Kansas City, Missouri
     64110,  EPA Contract No.  68-02-0228, April 1972.

5.   Schulz, E.J., Hillenbrand, L.J., Engdahl,  R.B., Source Sampling of Fluid
     Catalytic Cracking Plant (Electrostatic Precipitators and CO Boiler)  of
     Standard Oil of California, Richmond, California, Contract No. 68-02-0230,
     Task Order No. 3, Battelle, Columbus Laboratories,  Columbus, Ohio, July
     1972.

6.   Schulz, E.J., Hillenbrand, L.J., Engdahl,  R.B., Source Sampling of Fluid
     Catalytic Cracking, CO Boiler, and Electrostatic Precipitators at the
     Atlantic Richfield Company, Houston, Texas, Contract No. 68-02-0230,
     Task Order No. 3, Battelle, Columbus Laboratories,  Columbus, Ohio,
     July 1972.

7.   Anon.,  Compilation of Air Pollutant Emission Factors (Second Edition),
     U.S. Environmental Protection Agency, Office of Air and  Water Programs,
     Office  of Air Quality Planning and Standards, Research Triangle Park,
     North Carolina, April 1973.

8.   Cowherd, Chatten, Source Testing - EPA Task No. 6,  Standard  Oil Company,
     El Segundo, California,  Midwest Research Institute, Kansas City,
     Missouri, EPA Contract No. 68-02-0228.

9.   The Oil and Gas Journal, Annual Refining Survey, March 29, 1976.

10.  Ctvrtnicek, T., T. Hughes, C. Moscowitz, and D. Zanders, Monsanto
     Research Corporation, "Refinery Catalytic Cracker Regenerator SO .
     Control Process Survey," prepared for Office of Research and Development,
     EPA, Contract No. 68-02-1320, Task I, Phase I, September 1974.
                                     -91-

-------
11.  "New Energy Act May Hurt Refiners," The Oil and Gas Journal. May 17,
     1976.

12.  The Oil and Gas Journal, Construction Survey, 1976.

13.  "The Impact of Lead Additive Regulations on the Petroleum Refining
     Industry," by Arthur D. Little, EPA Contract 68-02-1332, Task Order 7,
     December 1975.

14.  "Striking Advances Show Up in Modern FCC Design," The Oil and Gas
     Journal. October 30, 1972.

15.  Ritter, R.E., J.J. Blazek, and D.N. Wallace, "Hydrotreating FCC Feed
     Could Be Profitable," The Oil and Gas Journal. October 14, 1974.

16.  "Complete Combustion of CO in Cracking Process," Chemical Engineering,
     November 24, 1975.

17.  Background Information for Proposed New Source Performance Standards,
     Vol. 2, Appendix:  Summaries of Test Data, EPA, Office of Air and Water
     Programs, Office of Air Quality Planning and Standards, Research
     Triangle Park, Publication No. APTD-1352b, June 1973.

18.  Mangebrauck, R.P., D.J. von Lehmdan, J.E. Meeker, "Sources of Poly-
     nuclear Hydrocarbons in the Atmosphere," USPHS Publication No. 999-AP-33,
     1967.

19.  Wollaston, E.G., "Sulfur Distribution in FCU Products, The Oil and Gas
     Journal, August 2, 1971.

20.  Huling, G.P., J.D. McKinney, and T.C. Readal, "Feed Sulfur Distribution
     in FCC Product," The Oil and Gas Journal, May 19, 1975.

21.  "Converting U.S. Refinery to Sour Crude Costly," The Oil and Gas
     Journal, April 9, 1973.

22.  "Shell Flue Gas Desulfurization Process Demonstration on Oil- and
     Coal-Fired Boilers," A.I.Ch.E. Meeting, Tulsa, Oklahoma, March 1974.

23.  Gary, J.H. and H.E. Scheweyer, "Crude Source & Desulfurization,"
     Petroleum Refinery, September 1953.

24.  "HDS & FCC Equals More Gasoline," The Oil and Gas Journal, May 17, 1976.

25.  EPA Standards of Performance for New Stationary Sources, Modification,
     Notification and Reconstruction, Federal Register, Vol. 40, No. 242,
     December 16, 1975.

26.  Ctvrtnicek et al, Monsanto Research Corporation, "Refinery Catalytic
     Cracker Regenerator SO  Control Process Survey," EPA Contract No.
     650/2-74-082, September 1974.
                                    -92-

-------
       APPENDIX A




SUMMARY OF EMISSION DATA
         -93-

-------
     The available emission data on uncontrolled and controlled fluid cata-
lytic cracker regenerators are presented in this appendix.  The uncontrolled
emission data are split into two categories according to principal pollu-
tants.  The first category contains emission data and test results for CO,
SOX, and NOX including the following:

     a)  V. H. Sussman Data (1957)

     b)  Summary of test data from EPA National Point Source Data Bank

     c)  Source Sampling Results on selected FCC units in SOCAL and ARCO
         refineries.

     In the second category are emission data which report hydrocarbons as
well as other constituents.  Included in this category are:

     a)  V. H. Sussman Data (1957)

     b)  Emission ranges reported in Ctvrtnicek, etal., "Refinery
         Catalytic Cracker Regenerator SOX Control Process Survey"

     c)  Typical uncontrolled FCC emissions presented in Exxon scrubber
         technical sales literature.

     Finally, typical emissions from the Exxon jet scrubber system applied
to a FCC unit are presented.

     Neither Standard of Indiana nor Exxon Company would provide hard data
on SOX or hydrocarbon emissions based on FCC stack sampling tests.  ADL
has informed both companies that EPA may request permission to perform on-
site tests to determine hydrocarbon emissions from UltraCat regenerator
(Amoco) and S0x/particulate emissions from the jet scrubber (Exxon).

     A sample calculation for estimating FCC regenerator flue gas volume
is also provided in this appendix.
                                     -94-

-------
    1.	SO. /NO, /CO
                                        TABLE  A-l

              EMISSIONS OF SULFUR OXIDES, AMMONIA,  AND CYANIDES FROM STACKS OF
           FLUID AND THERMOFOR CATALYTT CRACKING UNITSa (Sussman,  1957)
Type
FCC
FCC
FCC
FCC
TCC
TCC
FCC
TCC
TCC
TCC
TCC
TCC
TCC
FCC
TCC
SO3
Ib/hr
164
12.0
-1.20
8.90
1.25
-
6.90
5. 10
2.0
1.60
2.70
5.74
7.77
3.07
0.62
SOj,
Chemical anal.
Ib/hr
535
• 362
1,260
453
17.5
-
648
15. 1
14. 0
18.7
13.2
13. 0
11. 1
'205
24.4
ppm
438
512
2, 190
308
114
-
984
86
65
151
136
105
97
1, HO
141
MS,b
ppm
47
220
1.850
20
-
- -
-
15
10
-
91
-
60
360
15
'''ol ills
as SO^,
vol %
0.05S
0.540
0.220
0.031
0.011
-
0.098
0.011
0. OOS
0. 016
0. 016
0.015
0.015
o. no
0.01-1
Wl % SO,
in total
oxides
of sulfur
23.5
3.2
0. 1
1.8
6.7
-
1. 1
25. 0
n. o
7.9
17.0
30.6
41.2
1.4
2. 5
Nil {.
Ib/hr
130
27.0
20.5
26. 0
1.20
-
118
4.60
:». 40
2. JO
1.90
1. 56
3. 12
23. 0
i. KO
ppm
401
140
134
67
29
-
675
99
60
67
74
47
103
550
61
Cy.iiii
-------
I
VO
                                                     TABLE A-3


                                           TEST DATA FROM EPA DATA BANK
Company
Mobil
SoCal
Getty
Rock Island
Refinery
Ashland
Cities Service
Exxon
Cont. Oil
?
Exxon
Texaco
Exxon
Sun Oil
Sun Oil
Texaco
Gulf
Gulf
Chevron
Amoco
Location
Los Angeles
El Segundo
Philadelphia
Indianapolis
Kentucky
Lake Charles, La.
Louisiana
Montana
Montana
Montana
New Jersey
New Jersey
Ohio
Ohio
Tulsa
Philadelphia
Philadelphia
Utah
Utah
Approximate
Capacity
(Bbl/day)
50,000
70,000
50,000
15,000
60,000
7
?
16,000
9,000
30,000
40,000
130,000
53,000
70,000
31,000
20,000
52,000
6,000
153,000

Part.
33
100
15,300!
67
1,550
—
—
110
—
251
500
2,090
45
59
3
63
1,140
219"
176
tons/year
S0x , N0x HC
4,450 2,450 0
5,700
61,300
476 89 105
*
959
3,110 — 3
92 33
898
3,270
—
530 410 0
—
—
__
—
283 0
2,750 0
500
2,670
• 9
CO
0
—
—
33,700
6.0202
788
58,000
58,700
—
—
0
—
—
—
—
—
0
• —
—
1-
CO Boiler
or
Afterburner
Yes
Yes
Yes
No
Yes
7
Yes
No
No
Yes
No
Yes
No
No
No
No
No
Yes
Yes
         No  elect,  precipitator  or  cyclone.

         Corresponds to 99.8%  control efficiency.

-------
              RESEARCH REPORT
                    on
SOURCE SAMPLING OF FLUID CATALYTIC CRACKING
 PLANT (ELECTROSTATIC PRECIPITATORS AND CO
  BOILER) OF STANDARD OIL OF CALIFORNIA,
          RICHMOND, CALIFORNIA

         Contract No. 68-02-0230

            Task Order No. 3
         EPA Report No. 71-PC-20
                    to
     ENVIRONMENTAL PROTECTION AGENCY
          OFFICE OF AIR PROGRAMS
               July 6, 1972
                    by
     E. J. Schulz, L. J. Hillenbrand,
            and R. B. Engdahl
                BATTELLE
          Columbus Laboratories
             505 King Avenue
          Columbus, Ohio  43201
                   -97-

-------
VO
00
I
                                                       TABLE A-4

                                                   GAS COMPOSITION*
                                              Standard Oil of California
Run

No. Date Time
1 12-16-71 1

2
2 12-17-71 10

12
3 12-17-71 '2

4
:20

:40
:50

:00
:10

:00
Moisture,
Percent
(a)
pm
17.9
pm
am
19.4
pm
pm
19.7
pm
C02,
Percent
(b)
13.

14.
14.

14.
14.

14.
0

0
0

0
0

0
02. NOX,^
Percent ppm
(b)
4.

3.
4.

4.
4.

4.
(c)
0
104
5
0
100
0
0
116
0
02,
Percent
(d)
4.
3.
4.
__
--
4.
4.
4.
4.
0
7
5
_
-
0
0
0
0
CO,
ppm
(e)
14
14
14
0.5
0.5
0.5
0.5
0.5
0.5
«2.
Percent
(f)
83.

82.
82.

82.
82.

82.

0

5
0

0
0

0
               (a) EPA particulate train method (Federal Register, Vol. 36, Part  II, No.  159
                   (August 17, 1971).
               (b) Fyritc (grab sample).
               (c) Evacuated flask, EPA (grab sample).
               (d) Becknan Model 715 (continuous sample).
               (c) Bccknan Model 215A (continuous sample).
               (f) Calculated by difference.

                *  downstream of ESP and CO boiler

-------
SOURCE TESTING—KPA TASK NO. 8
     STANDARD OIL COMPANY
     Richmond, California
              by

          E.  P.  Shea
  Midwest Research Institute
 Kansas City,  Missouri  64110

 EPA Contract  No.  68-02-0228
  (Mill Project No. 3585-C)
             -99-

-------
o
o
                      DATE
                    2-08-73
                    2-09-72
                    2-10-72
                                               TABLE A-5




                                      SUMMARY OF  STACK  SAS RESULTSb
CO*/ NOX AS
(PPM-DRY) (L3/DSCF)
.0000143
10.0 .0000160
.0000121
.0000137
.0000149
9.0 .0000156
. .0000174
.0000184
.0000162
11.0 .0000157
,000017b
.0000189
N02 502
(PPM-DRY) (L3/OSCF) (PPM-DRY)
120 .0000483 292
135
101
116
125 .00006S8 397
131
146
155
153 .0000681 411
132
146
159
            aj  Determined with NDIR instrument and corrected  for C(>2 interference.



            b/  downstream of CO boiler.

-------
          SOURCE TESTING
          EPA TASK NO. 6
STANDARD OIL OF CALIFORNIA COMPANY
      El Segundo,  California


                by

          Chatten  Cowherd
    Midwest'Research Institute
   Kansas City, Missouri 64110

   EPA Contract No. 68-02-0228
     (MRI Project No. 3585-C)
              -101-

-------
                                            TABLE  A-6

                            SUMMARY OF RESULTS  (STACK GAS COMPOSITION



1
o
N>
1

CO*/
Run Date (ppm/vol)
1 12/14/71
2 12/15/71


3 12/16/71
4 12/15/71 5
NOX as
(ib/dscf)
1.31 x 10 ~5
1.39 x 10 ~5
1.47 x 10~5
1.24 x 10'5
1.46 x 10~5

1.72 x 10'5
1.97 x 10"5
1.67 x 10 ~5
1.75 x 10'5
1.64 x 10 -5
N02 S02
(ppm, dry) (ib/dscf)
108
114 2.23 x 10~5
121
102
120

142 5.48 x 10'5
162
138
144 6.09 x 10~5
135

(ppm, dry)
-
132


324
360
a/  ^xcludir.s Crsat analysis. — downstream of CO boiler.
b/  Deterrdned with KDJE instrxment and corrected for C00 interference.

-------
SOURCE TESTING—EPA TASK NO. 6
  ATLANTIC RICHFIELD COMPANY
    Wilmington, California
              by

          E. P. Shea
  Midwest Research Institute
 Kansas City, Missouri  64110

  EPA Contract No. 68-02-0228
   (MRI Project Uo.  3585-C)
             -103-

-------
                               TABLE A-7

                  SUMMARY OF RESULTS-(STACK GAS POLLUTANTS)1*

Run
IE

1W



2E



2W



3E



3W



Date (
1/11/72

1/11/72



1/12/72



1/12/72



1/13/72



1/13/72


COS/ NOX as
. prr.i . dr y ) ( Ib /ds c f )
35 1.87 x 10-5
2.04 x 10"5
6 2.00 x 10-5
2.04 x 10-5
2.33 x 10-5
1.91 x 10-5
12 2.09 x 10-5
1.90 x 10-5
1.97 x 10-5
1.80 x 10-5
12 1.14 x 10-5
1.52 x 10-5
1.51 x 10-5
1.61 x 10-5
12 1.92 x 10-5
2.02 x 10-5
2.78 x 10-5
2.40 x 10-5
37 1.78 x 10-5
2.00 x 10-5
2.17 x 10-5
KOp S02
(ppr\, dry) (Ib/dseJ1) (t>rr.n. dry)
157
172
168
172
19G
160
176 5.26 x 10-5 319
160
165
151
96 4.75 x 10-5 287
128
127
135
162 5.50 x 10-5 332
170
233
202
150 7.08 x 10-5 428
168
183
a/  Determined with NDIR instrument and corrected for C02 interference.
b/  downstream of CO boiler.
                                  -104-

-------
                RESEARCH Kni'OIlT
                       on
SOURCE SA.MPTJNn OP FT.mi) CATALYTIC  CRACKING,  CO
 BOILER, AND ELECTROSTATIC  ITJICIPITATOUS AT THE
   ATLANTIC RIC!lFIi:iD COMPANY,  HOUSTON,  TEXAS

            Contract No. 68-02-0230
                Task Order  No.  3
            EPA Report No.  71-PC-19

                       to
        ENVIRONMENTAL PROTECTION AGENCY
             OFFTCF. OF ATR  PROGRAMS
                  July 6,  1972
                       by
        E. J. Schulz, L. J. Hillenbrand,
               and R. B. Engdahl
                    BATTELLE
             Columbus Laboratories
               505 King Avenue
             Columbus, Ohio  63201
                     -105-

-------
                                    TABLE A-8
                                 GAS COMPOSITION
   Plant   Atlantic  Richfield  Company

an
3.
L


2


5


*




Percent
Moisture
Date Time (a)
12/8/71 4

6
12/9/71 11
12
2
12/9/71 4

5
12/10/71 11

1
:50

:00>
:55
:30
:40
:00

:35
:30

:10
pm
13.9
pm
am
pm 14 . 5
pm
pm
17.5
pm
am
16.5
pro
Percent
C02
(b)
13.
Percent NO. Percent
02
(b)
0
4.
0
ppm 02
(c) (d)
4


(8)
13.
13.
14.
13.
12.

12.
14.

14.
0
0
0
0
0

0
0

0
4.
3.
2.
3.
3.

3.
3.

3.
0
0
0
0
8

8
0

0
78 (g)
3
76 2.
3
3.
3.
89 3.
3.
3.
70 3.

2

8
8
8
0
0
0
CO
ppm
(c)
0
(8)
(8)
0
0
0
0
0
0
0
0
0
Percent
' (0
83

83
84

84
84.2

84.2
83

83
a)  EPA particulate train method (Federal Register,  December 23, 1971, Vol. 36, Part II,
   No. 247).
b)   Fyrite (grab sample).
c)   Evacuated flask, EPA (grab sample).
d)   Beckman Model 715 (continuous sample).
e)   Beckman Model 215A (continuous sample).
f)   Calculated by difference.
g)   Rain water in sample train.
* '  downstream of CO boiler.
                                        -106-

-------
 2.    Hydrocarbons.  etc.
                                             TABLE A-9
              TOTAL IIYDROCAKHON KMISSlOr :. FROM KLU1O AND Tl IKRMOKOK CATALYTIC
                             CRACKING UNITSIACKS1' (Sussman. 1957)

Type


FCC
FCC
FCC
FCCC
TCCe
TCCC
FCCd-e
TCCd
TCCd
TCCb,c

TCCb'c
TCC
TCC
FCC
TCC
Mass spectrometer

Hydrocarbons

7.4.
.. 3.1
2. 1
1
_
..
_
0. 4
0. 5

0. 1

0.5

0. 3
1. 4

Hydrocarbons

1,213
1, 150
760
98
_
_
_
308
•I.-IS-I

87

121

328
1.655

Wt 7o C and C

67.7
8-1. 1
68. '»
42. 3
_
_
_
•10. 9
r>5. 1

79.5

67. -1

51.2
61.9

Vol % C and C

87.4
9-1.6
85. 5
54. 1
_
.
.
70.8
SI. -1

77

67. 8

75. 3
18. 8
Infrared spec tropholorne lor
Hydrocarbons
(as hcxanc),
tons/day
2.80
0.89
0. 60
0. 30
0. 02
0. 02h
1.20
0.04
0. l'<
8

0.02
f
0. 02
-
0. 30
Hydrocarbons
(as hexane),
ppm
142
78
65
12
8

116
13
•n
-

14
.
9
Trace
108
aAll concentrations arc reported on a dry basis.
^Only the mass spoctromclr r rt-sulls for Units F-tiT and F-4T were riilialjlo.  Sinci- Units F-1T ancl F-2T
 and Units F-3T and F-4T aro twin unils, the data shown result from combining the twin units.
cNo methane present as determincfl hy mass spcrlrometer.
 Mass spectroniotcr determinations include oxygenated C^ and Cc hydrocarbons.
cThc mass  spcclromeler results were not reliable..
'The infrared spectrophotomeler results were not reliable.
^Concentrations of hydrocarbons arc below limit of accuracy of the  infrared spectrophoiometer.
 Infrared spectrophotometric determinations were made on Unit D-1T only. The results shown were
 obtained by assuming that twin Unit D-1T and D-2T  emit the same  quantity of hydrocarbons.
 Source:   Danielson, J.  D., Air Pollution  Engineering Manual.
                                               -107-

-------
                                   TABLE A-10

              EMICGTON  RAM'JF.3 FROM I'UJJI) CATALYTIC CRACKING  UNIT
                    REGENERATOR,  Ulil-'QHK AND AFTER CO BOILER
              Fresh Feed  Rate,  bpsd       156,000
              Recycle Feed  Rate,  bpsd     35,000
Stack Discharge Rate,
scfm (60°F, 1 atm, dry basis)
Temperature, °F
Emissions:
  Sulfur Dioxide, ppm**
  Nitrogen Oxides (as N02) ppm
  Carbon Monoxide, % vol.
  Carbon Dioxide, % vol.
  Oxyp;en, % vol.
  Moisture, % vol.
  Nitrogen, % vol.
  Hydrocarbons, ppm
  Ammonia,  ppm
  Aldehydes, ppm
  Cyanide:-., ppm
  Particulates, r,rains/scf
                                    Before
                                  CO Boiler
 484 ,000

1000-1200
 140-3300
   8-394
 7.2-12.0
10.5-11-3
 0..7-2.4
13-9 - 26.3
78.5-30.3
  98-1213
   0-675
   3-130
 0.19-0.94
 0.08-1.39
tt
           After
         CO Boiler*
      Up to 30?. Volume
      Increase
      (On Wet Basis)
          485-820
          Up to 2700f
          Up to 500 f
            0-14 ppm
           11.2-14.0
            2.0-6.4
           13.4-23.9
           82.0-84.2
          0.017-1.03
  *  Emissions after CO boiler will be  affected  by  the type of
     supplemental fuel and operating  conditions  in  the CO boiler
 **  It was reported that up to 60$ of  sulfur  oxides  in regen-  •
     erator flue gas may appear a:-. SO3  (see  paf.e 85)
  t  Estimated
  tt
     Sussman Data
  Source:  Ctvrtnicek, et al, Monsanto
         September 1974
                             -108-

-------
                           TABLE A-ll

         Typical Uncontrolled FCCU Emissions Compared  to
         	The Current Most Stringent Regulations	
                                  Typical Data
                              Before      After
                             CO Boiler  CO Boiler
Most Stringent
Existing Source
  Regulations
Federal
       (2)
Emission
Particulate (grains/DSCF)
Catalyst
Opacity (Ringelmann No.)
Gaseous Emissions
SO (vppm)
Jv
CO (vol.%)
NH., (vppm)
HCN (vppm)
NO (vppm)
Hydrocarbon (vppm)


0.2 - 0.6 0.1 - ().-'» 0.0118 O.lMSS
< 	 1.0 to 3.0 	 > 1-0 1.5

< — 200 to 700 	 > 13()-200(1) (3)
10 <0.1 0.02 0-O.S
200 < 10 * *
<200 
-------
3.   Jet Scrubber Emissions
                              TABLE A-12

                          EXXON JET SCRUBBER
                      Pilot Plant Performance Data
                        FCCU Stack Gas Scrubbing
Range of Pollutant
Loadings to Scrubbers
Performance (Collection Kfficioney)
of Scrubbers on Pollutants
SO
       200-500 vppm
SO
           95-99%
Particulate

  - Catalyst

       0.1-0.3 grains/SCF
  - Condensables
                **
       0.1-0.2 grains/SCF
Particular!-

  - Catalyst

           85-95%

  - Condensables**

       up to 901
Notes;

 *
  Typical SO  concentration is about 300 vppm.
**
  Sulfates and traces of hydrocarbons collected by wet  test method.
Source:  Exxon Engineering - Petroleum Department.
                                      -110-

-------
                                 TABLE A-13
                          SAMPLE CALCULATION OF GAS
                         VOLUME FOR FCC REGENERATORS
Note:  This calculation is for the case of a gas stream from a  traditional-
       type regenerator upstream of a CO boiler.

Assume coke composition is 96% C, 3% H, 1% S.
                              96
For 100 Ibs coke, there are:  y=- » 8 moles C


                               y * 3 moles H


                              •   " -03 moles S
           3 H + 8 C + .03 S + x 0  + 3.76 x N.	*•

                1.5 H20 + y CO + z C02 + .03 S02 + 3.76 x NZ


Assume that the ratio CO /CO is 1.0 (typical value).
Then, y - z - 4.0.  Thus, the equation can be rewritten:

           3H+8C+.03S + x02 + 3.76 x NZ 	•*

                1.5 H20 + .4 CO + 4 C02 + .03 S02 +  3.76 x NZ

To satisfy the oxygen requirement of the right-hand  side of this equation,

           x-^ + -|+4+ .03

             - 6.78

Therefore, the total moles of gas per 100 Ibs. coke  is:

           1.5 + 4 + 4 + .03 + (3.76)(6.78) = 35.02 moles

Assume the coke yield by traditional regeneration is 6.0% by wt. of fresh
feed.  Then, the weight of coke per barrel is as follows:


     (1 bbl)(0.9 sp. gr.)(8.33 Ibs/gal)(42 gal/bbl)(.06) = 18.89 Ibs coke/bbl

For a catalytic cracking unit processing 50,000 bbl/day:
                                \
     Gas volume = (35.02 moles/100 Ibs coke)(379 scf/mole)(18.89 Ibs coke/bbl)

                     (50.000 bbl/day)   Q, ...   ,
                     	'(24)(60)  y} - 87,055 scfm


                                    -111-

-------
   APPENDIX B
LIST OF CONTACTS
     -112-

-------
     In this Appendix are listed the names, addresses, and telephone
numbers of individuals contacted for information in the course of
this study,
VENDORS

     Richard C. Herout
     Coordinator, Environmental Engineering
     Universal Oil Products Company
     Process Division
     20 UOP Plaza
     Des Plaines, Illinois 60016
     (312) 391-2880

     Jim Montgomery
     Davison Chemical Company
     Division of W. R. Grace
     10 E. Baltimore Street
     Baltimore, Maryland 21202
     (301) 727-3900

     Albert Chatard
     Davison Chemical Company
     Division of W. R. Grace
     10 E. Baltimore Street
     Baltimore, Maryland 21202
     (301) 727-3900

     Chris Earl
     Davy Power Gas
     Drawer 5000
     Lakeland, Florida 33801
     (813) 646-7100
STATE AND LOCAL CONTROL AGENCIES

     Illinois - Dr. John Reed
                Supervisor of Permit Review Unit
                Illinois Environmental Protection Agency
                2200 Churchill Road
                Springfield, Illinois 62706
                (217) 782-5562

     Indiana  - Edward Stresino
                Chief of Enforcement
                Indiana Environmental Management Board
                1330 W. Michigan
                Indianapolis, Indiana 46206
                (317) 633-4273

                                  -113-

-------
STATE AND LOCAL CONTROL AGENCIES Continued
     Kansas
     Louisiana
     New Jersey  -
     Ohio
     Oklahoma
     Pennsylvania-
     Texas
Robert Wallace
Kansas Country Control Board
Kansas City, Kansas
(913) 321-4803

Orey Tanner
Administrator, Technical Assistance Unit
Louisiana Air Control Commission
P. 0. Box 60630
New Orleans, Louisiana 70130
(504) 527-5119

Tom Leonard
Department of Environmental Protection
Division of Air Pollution
P. 0. Box 1390
Trenton, New Jersey 08625
(609) 292-6724

Howard Johnson
Engineer
Ohio Environmental Protection Agency
361 E. Broad Street
Columbus, Ohio 43216
(614) 466-6116

John Drake
Air Quality Division
Oklahoma State Department of Health
3400 N. Eastern Avenue
Oklahoma City, Oklahoma 73105
(405) 271-5220

Frank Willard
Air Pollution and Noise Control
Pennsylvania Department of Environmental Resources
Norristown Regional Office
Norristown, Pennsylvania
(215) 631-2415

Dr. Frank Spuhler
Regulations Applications
Texas Air Control Board
1100 W. 49th Street
Austin, Texas 78756
(512) 451-5711
                               -114-

-------
STATE AND LOCAL CONTROL AGENCIES Continued

     Washington  -  Michael Landor
                    Air Quality Division
                    Washington Department of Ecology
                    P. 0. Box 829
                    Olympia, Washington 98504
                    (206) 753-2800
OPERATING COMPANIES

     Mr. R. C. Mallatt
     Manager, Environmental and Energy Conservation
     Standard Oil of Indiana
     200 E. Randolph Drive
     Chicago, Illinois 60601
     (312) 856-5485

     Charles T. Rau
     Environmental Control and Safety Division
     Exxon Research and Engineering Company
     P. 0. Box 101
     Florham Park, New Jersey 07932

     Terry Day
     Exxon Research and Engineering Company
     P. 0. Box 101
     Florham Park, New Jersey 07932
     (201) 474-6206
                                 -115-

-------
        APPENDIX C
REPORTS AND CORRESPONDENCE
           -116-

-------
     This appendix contains copies of written responses to our inquiries,
trip reports, and summaries of telephone calls made in connection with
this study.

1.   Visits

     As part of this study, ADL staff members visited with three companies,
Standard Oil of Indiana (AMOCO), Exxon Research and Development Company,
and Universal Oil Products Company.  The results of our discussions are
incorporated into this report.  Summaries of the meetings are presented
here.  In each case a list of questions was forwarded to the company prior
to our visit.

     a.  AMOCO
              REVISED NOTES OF MEETING BETWEEN A.D. LITTLE
                AND AMOCO OIL ON FLUID CATALYTIC CRACKING
     The following notes provide the revised answers on the specific
questions.

1)   Will HTR be universally adopted?

     Amoco feels that, in the near future, 1/3 of all FCC units will be
     equipped for HTR.  In the longer term, 1/2 will have this capability.
     This will be reached mostly through revamps of existing equipment.
     Of new units, the vast majority will be equipped for HTR.  Amoco has
     converted 5 of its 11 units and has plans for additional conversions.
     UltraCat technology has been licensed for use in four additional
     units.  In making the decision whether or not to revamp, the economics
     for each specific case must be considered.

2)   How much excess air is required for HTR?

     It depends on the catalyst and the regenerator design.  A range of
     3-25% excess air has been observed—for UltraCat, typically 5%.  When
     a CO promoter is used, more excess air is required.

     No changes in air blower capacity are needed since the reduced coke-
     make by HTR (better catalyst selectivity) cancels the additional air
     requirement per pound of carbon.  Air distribution becomes a critical
     factor, however.

3)   Under what conditions is a CO combustion promoter catalyst used?

     Promoter catalysts reduce the dilute phase temperature, and can thus
     be used where metallurgy is an important consideration.  There are
     yield penalties associated with promoter catalysts, however, due to
     the higher amount of carbon on catalyst.  In other words, regeneration
                                  -117-

-------
     with the promoter catalyst is not as effective as using a high tempera-
     ture, so catalyst selectivity suffers.  Amoco sees no unique benefits
     for these catalysts and has phased them out of its refineries.  Exper-
     imentation is continuing on a smaller scale, however.

A)   What is the mechanism by which promoter catalysts work?

     The promoter increases the rate of oxidation of CO without directly
     affecting the yield structure on the reactor side.  Amoco has observed
     no adverse environmental effects, such as the formation of metal car-
     bony Is .

5)   How do hydrocarbon and NO  emissions compare for HTR and CO boilers?
                              X

     From an HTR unit, the NOX is 5-20 ppm leaving the regenerator compared
     with >50 ppm leaving a CO boiler.  The differential is mainly attri-
     buted to the higher flame temperature in the CO boiler.  The nitrogen
     oxides for HTR appear to come from the air rather than from the nitro-
     gen tied with coke.

     Based on thermodynamic considerations, no significant amounts of
     residual hydrocarbons are expected to be present for the HTR.  At high
     regenerator temperatures and with excess oxygen, the rate of oxidation
     of hydrocarbons being higher than the rate of CO combustion is expected
     to result in insignificant amounts of hydrocarbon emissions.  A limited
     number of tests seems to confirm this.  These tests have shown that
     typically the hydrocarbon emissions for HTR are in the range of 3 to
     35 ppm.

     COS and HCN have not been analyzed for in the regenerator stack.
     However, from chemical equilibrium considerations, no COS and HCN
     are anticipated when there is excess oxygen.

6)   How difficult is it to revamp a unit for HTR and how expensive is it?

     Most FCC units can be revamped, but the economics vary a great deal.
     Size is not the determining economic factor.  Amoco feels that there
     is very little chance the revamp cost would exceed 50% of either the
     original or the replacement cost of the unit.

     Amoco says that revamping for HTR does not necessarily increase the
     capacity of a unit.  Whether the unit capacity increases or not depends
     on its operation prior to conversion to HTR.

7)   What are typical values for 803/802?

     The Sussman data (1/125 to 1/1) were quoted, and Amoco said simply
     that their data fall within this range.  In the 1971 NPRA Q&A session,
     it was pointed out that the ratio appears to depend on the metals on
     catalysts.  Amoco does see this trend.  It was pointed out that in
                                   -118-

-------
     a few cases they have seen ratios above those expected for equilibrium
     for the reaction:
                                 °2         r-  S°3
     at the regenerator temperature.  This is puzzling, and no theory was
     offered other than that the analytical techniques nay be poor.

     No increase in the ratio is observed when promoter catalysts are used.

8)   If NSPS on SOX were promulgated, would you meet it by hydrotreating
     the feed or by FGD?

     Amoco believes that hydrotreating is not a viable option due to the
     high level of hydrotreating required.  Even if a sulfur spec were
     placed on gasoline, Amoco would hydrotreat the gasoline fraction
     rather than the entire feed.

     Either FGD or feed hydrotreating is expensive.

9)   Are there any methods other than HTR and FGD for controlling SOX; e.g.,
     steam stripping?

     The Monsanto claims for steam stripping apply only to natural clay
     catalysts.  One does not observe the substitution of water for ad-
     sorbed sulfur compounds on zeolites.  Amoco further notes that the
     required steam stripping rates indicated by Monsanto are ten times
     higher than those typically used for FCC.  This represents a great
     expense .

10)  Is it possible that catalyst improvements will reduce SO ?

     Amoco is carrying out proprietary work in this area.

11)  Where do bottlenecks generally occur when a unit is expanded?

     Amoco did not feel any generalizations were possible.

12)  What are typical values for the equipment life and onstream factor?

     FCC units are usually down for about one month every 3-4 years.  This
     corresponds to an onstream factor of 97-98%.  Equipment life varies
     considerably.  Reactors and regenerators have lasted up to 30 years.

13)  What are your projections for FCC expansions?

     No projections are available.  Amoco is looking at all ways of debottle-
     necking its own units.
                                    -119-

-------
Other General Comments

     There was some discussion of the correlation between feed sulfur and
coke sulfur.  Two publications—one by Amoco and the other by Gulf—have
addressed the issue; there was poor agreement between the two sets of data.
Vasalos claims that the difference lies in the fact that Gulf used hydro-
treated feeds for most of its data.  If the points are taken for virgin
stocks only, the two articles agree fairly closely.  For unhydrotreated
feedstocks, there appears to be a correlation between thiophenic sulfur and
coke sulfur.  Since thiophenic sulfur is difficult to remove by hydrotreat-
ing, severe hydrotreatment of feeds is necessary to reduce coke sulfur
levels.

     One would expect, generally, that SO  emissions will go up as the FCC
feed sulfur increases.  As sour crudes are used increasingly, SO  emissions
will therefore increase.

     b.  Exxon Research and Engineering Company


                  EXXON'S RESPONSES TO ADL QUESTIONNAIRE


A.   CHARACTERIZATION OF FCC REGENERATOR EMISSIONS

     1.  Q.  Is there a correlation between FCC feed sulfur and catalyst
             coke sulfur content?

         A.  Studies by American Oil Co. indicate that the sulfur levels
             in coke are proportional to the feed sulfur level.  Their
             findings indicate that the weight percent sulfur on coke
             produced from virgin feeds is about 0.7 of the weight percent
             sulfur in the feed.  No definite correlation between fuel
             sulfur compound type and coke sulfur level has yet been
             determined.  This is due to the large number of unanalyzed
             sulfur compounds in the feed.

     2.  Q.  How do other operating variables such as cracking severity,
             recycle ratio, regenerator operation affect sulfur emissions?

         A.  Lowering of cracking severity and recycle ratios will result
             in a reduction in the quantity of SOX emissions which is pro-
             portional to the reduction in total reactor coke.  High tem-
             perature regeneration, however, reduces SOX concentrations
             somewhat more than what would normally be observed for a given
             reduction in reactor coke make.
                                    -120-

-------
3.  Q.  What is the range of recycle ratios commonly used?  NPRA
        (Blazek) stated that the nationwide average in the U.S. is
        167.  of F.F.

    A.  Recycle ratios are dependent on the type of catalyst a refiner
        is using.  Recycle ratios from 0 to 15 vol% on F.F. are common
        for operations utilizing a zeolitic catalyst.  Operations
        with amorphous catalysts can be as high as 100 volZ on F.F.
        Recycle ratios will also vary seasonally as refiners gear
        up to produce either heating oil or gasoline products to
        meet consumer demands.

4.  Q.  What is a typical 803/802 ratio for regenerator effluent?

    A.  Typical S03/S02 ratios for CO boiler effluents, as operated
        by Exxon, are about 0.03 or less.

5.  Q.  If SOX NSPS standards were promulgated, would you meet them
        .via flue gas desulfurization (FGD) or feed hydrotreating (HT)?

    A.  It is not possible to state generally which process, FGD or
        feed hydrotreating, is preferred in meeting SOX standards.
        For a particular refinery the preferred system would be based
        primarily on economic considerations.

6.  Q.  Do you know of any means other than FGD or HT for SOX control?
        For example, steam stripping?

    A.  Presently,  our only methods for SOX control are by flue gas
        desulfurization and feed hydrotreating.  However, we are
        aware that  steam stripping is being investigated on a
        laboratory  scale elsewhere.

7.  Q.  Could any catalyst improvements be made to reduce SOX?  It  is
        reported that silica-magnesia amorphous catalyst could reduce
        regenerator SOX emissions.  (NPRA-Blazek)

    A.  We presently do not know of any catalyst or catalyst improve-
        ments that  would result in the reduction of SOX regenerator
        emissions.

8.  Q.  Will high temperature regeneration (HTR) be universally adopted?

    A.  Exxon U.S.A. units do not operate under HTR for CO control
        since they  all have existing CO boilers.  However units
        are being converted to HTR due to the economic incentive
        of incremental yields.
                             -121-

-------
 9.  Q.  How much excess air is required  for  HTR  to  ensure  essentially
         complete CO combustion?

     A.  The quantity of excess air required  to ensure  complete  CO
         combustion is dependent on catalyst  types used.  Combustion
         promoting catalysts require approximately 0.6  percent excess
         air, while ordinary zeolitic catalyst need  approximately 1.5
         percent excess air for complete  CO combustion  in HTR.

10.  Q.  Under what conditions are CO combustion  promoter catalysts
         used?

     A.  CO combustion promoting catalysts are generally used when
         it is not economical to use CO boilers or high temperature
         regeneration to meet desired CO  levels in flue gases.

11.  Q.  What is the general mechanism by which the  new catalysts
         work to promote CO oxidation? Are these catalysts essential
         for HTR?

     A.  A possible mechanism by which the new catalysts works is
         that a finely dispersed layer of noble metals  on the catalyst
         surface increases its ability to oxidize CO.   These catalysts
         are not necessary for HTR operation.

12.  Q.  How do hydrocarbon and NOX emissions from HTR compare  with
         emissions with a CO boiler?  'Can you provide test  results
         on emissions from FCCU?

     A.  There is no significant difference  in hydrocarbon  emissions
         from HTR or a CO boiler.  Concentrations of less  than  10 vppm
         hydrocarbon have been observed in both  types of  operation in
         Exxon units.  NOX emissions from either  configuration  are
         lower than 200 vppm.

13.  Q.  How difficult is it, in general, to  revamp  a FCCU  for  HTR?
         How expensive is it?

     A.  Revamping an FCCU for HTR operation  involves the replacement
         of all carbon steel regenerator  components  with  a  more  heat
         and erosion resistant material such  as a type  304  stainless
         steel.  Conversion to HTR can be accomplished  in a normal
         turnaround.  A recent conversion of  a 45 kB/D unit to HTR as
         estimated to cost 1.8 M$.  This  cost includes  replacement of
         cyclones, plenum chamber, cyclone diplegs,  regenerator  grid
         and seals, and the catalyst overflow well.
                             -122-

-------
 14.  Q.  If FCCU expansion is necessary, where are the bottlenecks, and
         what is done to increase the unit's capacity?  What effect do
         these changes have on emissions?

     A.  In typical FCCU expansions, the major bottleneck areas are
         in the overhead gas compressors, solids separation systems,
         air blowers and feed preheat trains.  Fresh feed capacity
         increases are sometimes obtained by using more active catalysts
         and more severe operating conditions to reduce the amount of
         recycle to the reactor.  SOX and CO emissions would be reduced
         if the regenerator were converted from low temperature regen-
         eration to HTR in an expansion.  Otherwise flue gas emissions
         will be proportional to coke make.

16.  Q.  What is the average equipment life for a FCCU?  Average
         stream factor?

     A.  The life of specific equipment  components  varies  a  great  deal,
         however, the average economic  life of a  FCCU  is 16  years.
         Average stream factors for  domestic units  are 967..

17.  Q.  Are ttere any available projections of FCC capacity expansions?
         How would FCC feed desulfurization trends  affect  these projec-
         tions,  i.e., yield improvements?

     A.  We do not have available any such projections.
                        EXXON SCRUBBER TECHNOLOGY

 1.  Q.  Is there any literature on application of this control technology
         to FCC regeneration?

     A.  The  technology is available for licensing and the pertinent
         brochure is available.

 2.  Q.  Has  the control  system been applied to FCCU?  Where are these
         installations?   How long have they been in operation on this
         source?

     A.  LOCATION                     SIZE  (Design)       STARTUP DATE
                                         (ACFM)

         Baytown, Texas                300,000            March, 1974
         Baytown, Texas                510,000            May,  1975
         Baton Rouge,  Louisiana        860,000            February, 1976
         Bayway, New Jersey            590,000            May,  1976
                                 -123-

-------
3.  Q.  Were there any specific problems at the above locations?  Upset
        conditions?  Problems during start-up and shutdown?  What was
        total time available to the FCC units since start-up?

    A.  • Overall performance has been good.  There have been some metal
          loss problems in the scrubbing liquid delivery system due to
          erosion and corrosion.

            + The reasons for the problems have been identified and
              solutions are being implemented.

        • Service factors on the two Baytown scrubbers (see answer to #2)
          have been 94 and 93% respectively since the beginning of 1975.

4.  Q.  What SO  and particulate (if applicable) removal efficiency is
        attainable with the control system?  Are there any supporting
        data available?  Can we visit or phone contact the operating
        companies?

    A.  • Pilot plant operations have demonstrated the following:

            - SO  removal efficiencies of 90+%

           _- Particulate collection efficiencies in excess of 85%

        • Initial commercial operations have demonstrated the following:

            - SOX removal efficiencies in excess of 90%

            - Particulate collection efficiencies of up to 90%


        • Testing of the Baytown scrubbers have confirmed that they have
          met all local control regulations and they also have met the
          Federal New Source Performance Standards for dry particulates.

        • Supporting data has been made available through a licensing
          brochure and compliance data have been sent to local control
          authorities.

        • Any follow-up should be handled through Exxon Research and
          Engineering Company.

5.  Q.  Are these problems related to oxidation and presence of 803?
        What is absorption capacity  for NOX?  Do installations have
        pre-scrubbers for removal of particulate matter?

    A.  • While there are no problems with 803 being present, it does,
          however, tie up part of the active component of the system's
          buffering agents.
                               -124-

-------
        • S03 levels are low and since the active portion of the buffer
          itself is subject to partial auto oxidation, no performance or
          operability debits occur.

        • To date the system has not been tested for NOX removal capabilities,
          Such testing would only take place once all commercial systems
          have completed their shakedown phase.

        • The Exxon scrubbing system is capable of removing both parti-
          culates and sulfur oxides in a single operation.  None of the
          Exxon Co. U.S.A. installations have particulate removal
          devices upstream of the scrubber.

        • The single step removal of both particulates and SOX is a major
          processing advantage for our scrubber system.

6.  Q.  For new plants, what kind of guarantee would be extended?  SC>2
        removal efficiency?  Outlet concentrations of SC>2 (ppm SC>2 on
        dry gas)?  Operating factor?  What are important operating
        conditions for the guarantees to apply?  What is the time period
        for guarantees?

    A.  • Typically scrubber collection efficiencies are

            •*• Particulate -- 80+% collection efficiency
            + SOX — 90+% collection efficiency

        • However, in general, we are still firming up our guarantee
          package.

        • Many of the factors which would go into the guarantee package
          would be subject to specific negotiations for the particular
          location and application.

        • Our present thinking on the guarantee package is as  follows:

            - We would guarantee the level of particulate and  SOX in
              the outlet gases from the scrubber for a specific design
              basis for the inlet gases and particulate to the  scrubber
              at designed operating conditions.

        • However, depending on specific conditions an outlet  "stopper"
          (vppm of SOX) may be applied.

            + Scrubber "service factor" is highly dependent  on  the
              owner's operating and servicing philosophy for the unit
              and is not usually covered by guarantees.

            + Once our guarantees were met, we would normally  consider
              this successful demonstration of the unit and  allow no
              additional time factor.
                                  -125-

-------
7.  Q.  Control system experience in refinery applications.  How many
        plants in the U.S.?  Europe?  Japan?  Operating history?  How
        does experience in Europe and Japan differ from U.S. practice?

    A.  • Exxon Co. U.S.A. currently has three (3) operating scrubbers
          and one (1) scheduled for startup within this quarter.

        • No units are currently operating or under construction outside
          of the U.S.A.

        • Capacity of the operating units is as follows:

            - 300 kACFM Baytown, Texas FCCU #2 Scrubber
            - 510 kACFM Baytown, Texas FCCU #3 Scrubber
            - 860 kACFM Baton Rouge, La. PICA #2/PLCA #3 Scrubber

        • Slated for startup within this quarter is the Bayway, N.J.
          FCBW #2 scrubber with a capacity of 590 kACFM.

        • By mid 1976, 2260 kACFM of the flue gas will be treated by
          the Exxon scrubbing process.

        • Operating experience has been covered in response to question
          No. 3.

8.  Q.  What, in your opinion, are the major competing processes?  What
        are the advantages and disadvantages of the competitive system?

    A.  • There are possibly two systems which are competitive with
          Exxon's Jet-Ejector venturi scrubber for both particulate
          and SO,, removal.  These are:
                A

            -  Combined FCCU feed hydrodesulfurization (HDS) with ESP's
            -  High energy venturi scrubbing


        • HDS/ESP's in general have the following advantages/disadvantages

            - Advantages

                + FCCU yield credits through HDS

            - Disadvantages

                + Operation of ESP sensitive to changes in FCCU operation
                +  Large onsite plot space requirements

        • High energy venturi scrubbing in general has the following
          advantages/disadvantages.

            - Advantages
                                -126-

-------
               4 Lower liquid circulation rate and consequently smaller
                 circulating pumps.

           - Disadvantages

               4 Is only viable if  "free" gas pressure drop is available
                 at a suitable location such as downstream of a CO boiler.

               4 If fans and/or blowers are required maintenance is
                 increased.

               4 Investment is greater than jet ejector scrubber when
                 fans and/or blowers are used.

       •  In general, we have found  the jet ejector wet gas scrubber to
          have  the following advantages controlling both participate and
          sulfur oxides:

           - Control of several pollutants
           - Operation is relatively insensitive to changes in -FCCU operation.
           - Excellent onsite plot  utilization

               4 Less duct work
               4 Single unit
               4 Offsite storage and disposal

           - Technology will not readily become outdated.

           - High service factor, low maintenance costs.

           - Investment and operating costs are lower than comparable
             schemes for simultaneous control of particulates and sulfur
             oxides.

       •  Conclusions reached as to  advantages/disadvantages apply in
          general.  Specific refinery conditions such as environmental
          regulations; ploc space; "grass-roots" or existing units;
          processing requirements; and other factors vary  the choice
          of the optimum emission control system from refinery to refinery.
          Thus, the choice of system is highly dependent on the specific
          refinery.

9.  Q.  Do FCC  units  represent  a  significant market  for  air pollution
        control systems?

    A.  • There are  approximately  180  FCCU's  in  the  USA

        • Outside  Exxon Co. USA we cannot  speak  with  any authority on
          what  percentage of  units would  require additional control
          measures.

        • While there  appears  to  be  a  potential  market  information as  to
          the  true  potential  must  come  from  the  operating companies.


                                    -127-

-------
                            ADL MEETING NOTES

                   EXXON R AND E--FLORHAM PARK. N.J.

     Dr. Jashnanl and Mr. Stickles of Arthur D. Little (ADL) attended a
meeting at Exxon Research and Engineering (ER & E) in Florham Park on
April 23, 1976.  The meeting was held at the request of ADL to discuss
air-borne emissions from FCCU and ER & E's control technology for FCC
regenerator emissions.  Representatives from Exxon R and E included:

         John Cunic           -  Project Engineer
         Terry Day            -  Senior Project Engineer
         Charles Rau          -  Project Engineer
         Herbert Schroeder    -  Technology Sales Engineer

     Prior to the meeting, ADL had submitted a list of discussion questions
against which ER & E had prepared written answers.  A copy of ER & E's
Q and A write-up is attached.  The discussion of each Q and A during the
meeting elicited other informative responses which are presented by
question nos. below.

A.   CHARACTERIZATION OF FCC REGENERATOR EMISSIONS

Q-l  Copies of the referenced paper were provided by ER & E.  ER & E are
     studying ways to predict coke sulfur from feed composition, but
     accurate modeling is difficult.  Process design of FCCU is based
     upon operating experience and FCC mass balance data.

     ER & E believe the trend is toward higher sulfur in feedstocks.
     Arabian Light seems to predominate ad a marginal crude, however,
     Arabian Heavy should be considered as a study case.  Typically, FCC
     feed from Arabian Light run 2-3% sulfur.

Q-2  With conventional regeneration, sulfur emissions are directly
     proportional to coke make.  For HTR, the incremental reduction in
     sulfur emissions is slightly better than 1 to 1 with reductions in
     coke make.  The reason for this is not known, but it is observed.

     Because HTR removes catalyst coke more effectively, the regenerated
     catalyst has a higher activity than with conventional regeneration.
     Hence, for a given yield structure, about 25 percent less coke is
     made.  This means that feed or severity can be increased until the
     practical coke make limit is again reached.  This potential for
     yield or capacity improvement is the major attractiveness of HTR.

     The reduction of CO by HTR to 500 vppm has been demonstrated,
     however the temperature required (^1400°F) to achieve this CO
     level is not necessarily the economic optimum for yield improvement.
     ER & E believe there are temperature levels below that needed for
     compliance with CO regulations that give economic operation.  At
                                  -128-

-------
      these conditions CO boilers would still be needed to control CO
      emissions.  In conclusion, the main incentive for HTR is yield
      improvement.

      Catalyst addition with HTR may be slightly higher to maintain
      activity. However, catalyst attrition is nearly the same with HTR
      as practiced  by Exxon USA.

 Q-3 and 4    No significant additions to stated answers.

 Q-5  There are no  sweeping generalities, each situation has to be
      individually  evaluated.  In fact, ER & E look at both alternatives
      (GO-fining vs. scrubbers) when evaluating emission controls for
      FCCU.  Naturally, plot space is an important factor in retrofit
      situations.

 Q-6  Beside Monsanto's interest in steam stripping, ER & E is not aware
      of any other  developments.

 Q-7  Answer as stated.

 Q-8  For new FCCU, ER & E believe some form of HTR would be applied,
      primarily to  improve yields.  Whether to go to CO control by this
      means would depend on the relative importance of steam (CO boiler
      economics) compared to the cost of HTR.  HTR is the least capital
      intensive method of control; upgrading of metallurgy being the
      principal cost.

      For existing  FCCU with CO boilers, there is no incentive to go all
      the way to control CO by HTR.  In this case, optimum conditions for
      yield gains would be selected.

 Q-9  Exxon USA does not have any CO combustion promoting catalysts in
      operation.

Q-10  Promoter catalyst permit CO combustion to proceed at lower
      temperatures.  This is of importance with existing FCCU where
      metallurgy limitations might not allow CO control by unpromoted HTR.
      However, to get yield improvements associated with HTR, high
      temperatures  are essential.  Consequently, promoter catalysts are
      used where CO control is the overriding consideration (short term
      solution).  The ability to meet the CO limit at conventional
      regeneration  temperatures will depend on the effectiveness of the
      promoter catalyst.

Q-ll  Answer as stated.

Q-12  The concentration levels in both cases are after the CO boiler.
      Due to the difficulty of sampling upstream of the CO boiler, ER & E
      do not have a lot of test measurements on emissions leaving the
      regenerator.   Reproducibility is also difficult due to the low
                                   -129-

-------
      concentration of NOX and hydrocarbons.  Charles Rau will check on
      the availability of test data and determine if it can be released
      to ADL.

Q-13 and 14    In the past, FCC capacity was increased by 1) the use of
      improved zeolite catalysts and 2) Increasing flow rate by raising
      operating pressure.  HTR is the latest innovation for expanding FCC
      capacity.  The reduction in coke make over conventional operation
      is utilized to increase capacity by running backup to the coke
      limit.  Capacity can be increased by feed pretreatment, but this is
      usually not economical.

      In general most operators have taken advantage of the expansion
      opportunities available through catalyst improvement and pressure
      operation.  Immediate projects do not suggest that another round of
      expansions based on catalyst improvements is imminent.

      The cost of conversion to HTR depends on the metallurgy of the
      existing unit.  A typical cost is given in answer 13.  ADL will
      find out what criteria are applied to define a new source and
      whether an increment in capital investment is included.

Q-15  FCC is a mature technology and the basic conversion, and, although
      design differences exist, yields and operating requirements are
      about the same for the licensed processes.  For a given feedstock
      and operating condition, gaseous emissions should be basically
      similar.  Particulate emissions will depend on cyclone design and
      the number of stages employed.

Q-16  ER & E see some evidence of a "new car fever" in FCCU expansion
      planning.  Consideration is being given to new units, especially in
      cases where the existing unit is of World War II vintage.  The
      ability to comply with local emission standards will be a factor
      affecting such decisions.

Q-17  ER & E suggested that ADL contact API, NPRA and the editorial staffs
      of 0 & GJ and Hydrocarbon Processing for projections on FCC capacity
      additions.
 B.   EXXON SCRUBBER TECHNOLOGY

 Q-l  Answer as stated.

 Q-2  The Baton Rouge scrubber handles gas from two FCC units.

 Q-3  The problems were associated with the materials of construction.
      Better materials are now being specified.

 Q-4  The ER & E scrubber system has been approved by the Texas Air Quality
      Board and can meet the state particulate and primary ambient air
      quality standards.


                                  -130-

-------
Q-5  No additions.


Q-6  No additions.


Q-7  Typical liquid to gas ratios for ejector scrubbers are 40-120 gpm/
     1000 acfm.

Q-8 and 9    No additions.
     As a final note, it was pointed out that the auxiliary fuel used in
CO boilers can increase the sulfur concentration in the flue gas from
FCC units.
     C.  UNIVERSAL OIL PRODUCTS COMPANY


                            UOP MEETING NOTES
                             April  29, 1976





Those present were:      Peter Stickles                   - ADL

                        Leigh Short                      - UMass


                        Hal Hammershaimb                 - U.O.P.

                        Dick Herout                      - U.O.P.


                        Dennis Michaelis                 - U.O.P.

                        Dick Conser (pt.  time)           - U.O.P.



The meeting began with introductory remarks by Peter Stickles, outlining the


objectives of the study, and emphasizing that no proprietary information was

being sought.  UOP commented that it was their understanding that this study

would also not get into costs or technology under development.  Aside from

these areas they intend to cooperate fully but pointed out that since they are

not an "operating" company they just do not have access to and cannot furnish


some of the data questioned.
                                  -131-

-------
The numbers  below refer  to  the  question numbers as  sent  to  U.O.P.









Al.   There is  no  simple  correlation  between  FCC feed  sulfur and  coke sulfur.




A  formerly widely used rule of  thumb was  that  feed  sulfur equals coke sulfur,




but this  "rule" is not too  accurate.   U.O.P. has  done limited  commercial




testing for  flue  gas  sulfur oxides—there has  been  no need  for extensive




commercial testing.   Pilot  plant  data, while accurate in the analytical




sense, must  be used with care because these  units are not always in heat




balance.  This may affect the sulfur balance.  UOP  agreed that the Gulf API




paper of  May 1975 was probably  the most complete  pilot plant information in




the literature.    UOP also  emphasized that the quantity  of  sulfur in the coke




is feed specific.









A2.  The effect of other variables on sulfur emissions is hard to determine




—very little commercial data is available.  At high severity,  if the sulfur




compounds crack, the sulfur will end up as additional l^S.  Refractory sulfur




compounds will tend to concentrate in the cycle oils.   As recycle increases, it




might be expected that more sulfur will show up on the coke.  The residual sulfur




on the coke after regeneration is very low.









A3.  UOP plants in the U.S.  probably average less than 16% recycle.  New plants,




using zeolite catalysts,  can be designed for 5% or lower.  Older plants may have




originally been designed at 40% or even higher.   A minimum recycle flow is




generally used for return of entrained catalyst.   The nationwide average will




probably continue down.   Recycle ratios are not much different for HTR operation




—if anything the ratio is  slightly less.
                                   -132-

-------
A4.  Data is available indicating up to about 10% of the coke sulfur goes to




S03, but UOP pointed out that they have some questions about the analytical




method used for 803.  (They used a modified version of the LA method rather




than EPA's method.)  U.O.P. knows of no data comparing S03 concentrations with




varying regeneration techniques.










A5.  There is no single, simple answer to this question.  UOP stress that any




 standard-levels considered should be high enough so as to not take away control




 options that are viable and economic.  If standards are developed (UOP questions




 their necessity and value at this time), they should be of such structure that




 several control options, including flue gas desulfurization and feed hydrotreating,




 are possible alternatives depending upon a refiner's specific considerations




 relating to such things as charge stock, product requirements, etc.  Both primary





 alternatives will  effectively  reduce SOX emissions and have distinct benefits.




 FGD may be less costly, depending on the process employed, but in the case




 of feed hydrotreating, it should be noted that there may be significant




 benefits in terms  of increased yields, reduced gasoline and fuel oil sulfur




 levels and the potential for increased unit throughput.  The choice for a




 particular case would depend on these considerations as well as individual




 economic factors,  as discussed in UOP's 1974 NPRA paper.









 A6.  UOP is not aware of commercial employment of means other than flue gas




 desulfurization or feed hydrotreating for FCC SOX control.  They are aware




 of the Monsanto report which suggested steam stripping.  Tho UOP does not




 have a concrete opinion on this, they believe that the industry feeling




 is that this would be a questionable method.
                                     -133-

-------
A7.  Catalysts could probably be developed to reduce SOX emissions to some




extent, but U.O.P. does not specifically do this type of research.  The catalyst




manufacturers may have comments in this area.








A8.  There is no single answer.  HTR is not for everyone.  These refineries




who have CO boilers may not be able to justify HTR—in any event the CO boiler




would likely be retained for some heat recovery.  The refinery steam situation




is important to this economic evaluation.  For new cat cracker units, HTR with




and without a CO boiler are possible options.  Older units without CO boilers




are considering revamping for HTR to comply with emission standards as well




as to improve energy conservation.  In addition to these factors, HTR has an




advantage in improved activity of regenerated catalyst due to lower levels of




residual carbon on regenerated catalyst.  See UOP 1976 NPRA paper.










A9.  The amount of excess air required depends upon operating conditions and




design.  Very little excess air is needed if the conditions are optimum.  The




typical range is 0-2% oxygen.  With conventional regeneration, the range is




0.1-0.2% oxygen.   U.O.P.  has data on the effect of excess air on CO but not




SOX.   Too much air may not be beneficial due to temperature and residence time




considerations.









A10.   CO promoter catalysts do not change the dense phase temperature,  but




do lower the peak temperature in the dilute phase.   The catalysts cost more but




can save money by requiring fewer metallurgical changes during a revamp, i.e.,




increases the degree of flexibility.  There are no yield changes (using promoter
                                  -134-

-------
catalyst) per se.  U.O.P. does see widespread use of these catalysts.  For a




light feedstock, requiring low regeneration temperatures, promoter catalysts




are more likely to be used.









All.  The promoter catalysts are not essential for HTR operation.








A12.  U.O.P. does not have data showing a comparison of HC/NOx from HTR versus




a CO boiler.  Intuitively one would expect low levels of hydrocarbon in both




cases since CO destruction is usually rate limiting, but lower NOX for HTR




(since no CO boiler).  Metals are not present in significant quantity in the




fines emitted to the atmosphere—cyclones remove 99% of the particulates, ESP




almost all the rest.










 A13.   Each unit must be looked at individually.   Often,  for old  units,  the




 carbon steel in the regenerator,  cyclones,  slide valves,  etc.  will have to be




 replaced by 5 chrome or 18-8 stainless (which is better).   The air blower is




 often the bottleneck.






       ESP resistivity changes are not a problem with HTR if the  gas is  cooled




 to the "old" temperature—about 700°F.






       Regarding costs,  UOP pointed out that it is not reasonable to tie NSPS




 "Reconstruction" applicability for revamps to the cost of the regenerator alone.




 The way the regulations now read, what is actually a moderate revamp of an FCC




would likely fall into the "Reconstruction" category of the regulations,  which




 require a refiner to comply with the same stringent and costly standards as
                                   -135-

-------
new units.  This could frighten some refiners away from otherwise justified




revamps which would have reduced CO emissions and conserved energy.








A14.  Expansions can be done in a number of ways.  Alternatives or changes




include:





      a)  Zeolite catalysts (most FCC units now have this)




      b)  Raise the pressure—this alters the entrainment of catalyst to




          the fractionator, but does increase capacity




      c)  Desulfurize the feed—this can help if the process is air limited




      d)  Install a parallel fractionator—some refineries have done this




          plus changed the air rate.





      Refineries cannot tolerate high particulate emission rates, because of




the expense of replacing the catalyst lost.











A15.  ADL clarified the question by pointing out that they were not inquiring as




to competitive differences or advantages between the processes, but rather whether




they all would fall into the same emission category.  In this context,  UOP said




that there are probably only minor differences in emissions from the several




processes, but there are design differences.









A16.  The operating factors are 95% (average 7 days/year downtime).  Some FCC's




have run continuously for six years, before maintenance turnarounds.  SO  standards




would have to account for the long run times of FCC units versus the shorter




times of FGD units.
                                   -136-

-------
A17.  U.O.P. projects FCC operating capacity growth rate to 1985 of 0.5% per




year (U.S. plus Canada).  This reflects some change in lifestyle due to higher




energy costs.  If there is a severe adjustment in lifestyle—they project virtually




no growth rate in FCC capacity to 1985.  It is expected that in the U.S. 20,000 B/D




is about the minimum size FCC unit that will be installed.  U.O.P. projects an




increase of roughly 150,000 B/D FCC capacity by 1980 above the 1974 capacity of




approximately 5.2 x 10*> B/D.  This figure includes revamps.  They had no information




on how to split new versus replacement or revamp capacity increases.  By comparison




the FCC growth in the period 1966-74 was 2% per year.









FGD TECHNOLOGY




The following points were brought out by U.O.P.





1.  SFGD capital costs  should be less than VGO desulfurization (VGO also uses




    more hydrogen than  SFGD, to achieve the same flue gas sulfur level.)






2.  S02 removed  (recovered) from the FCC will go to the Glaus unit.  Generally





    no major changes  are expected here, as heat  removal  (steam generation)




    following H2S oxidation is  the usual  thruput limitation  on a Glaus  plant




    while  S02 into  a  Glaus  may  bypass  this stage.  The Glaus plant  should  be




    checked in  each case.






 3.  In establishing VGO-desulfurization severity,  flue gas SOX standards are




    not necessarily controlling.  Product specifications on  sulfur may  establish




    the degree  of sulfur removal in  the VGO desulfurization  unit.






 4.  There  are no commercial Shell/UOP  FGD units  installed  in the U.S.   There




    is one in Japan processing  flue  gas from  a residual  fuel oil-fired  boiler




     (40 MW).





                                     -137-

-------
5.  UOP can provide no information on guarantees between JST  (Japan-Shell  Tech.




    Co.) and SYS  (Showa Yokkaichi Sekiyu).  UOP is prepared to guarantee flue




    gas capacity, percent desulfurization and acceptor life and perhaps other




    factors depending upon the specifics of a particular application.  Guarantees




    can be fully  discussed only in view of the specific technical and  economic




    factors and financial liabilities.









6.  The reactor internals of the SFGD reactor are designed to accept high  loadings




    of entrained  solids and can be located in the flue gas train without regard




    to solids removal equipment.  The pilot operation at Tampa Electric has




    confirmed the above.






7.  NO can be removed to any desired level.  Removal to 100 ppm can be achieved  at




    modest incremental cost increase over that for desulfurization.  Performance




    data on the SYS unit are available.







8.  It is considered that COS passes through the SFGD unit unreacted.   This can




    be checked in the laboratory if necessary.








9.  Above 90% sulfur removal,  the SFGD costs starts to rise significantly.




    There is no specific technical limit.








10. Since the acceptance cycle of SFGD is an oxidation,  it is necessary that the




    feed gases are in a completely oxidized condition and that a  small amount of




    oxygen is present.  SFGD could not be applied to flue gases from an uncon-




    trolled FCC unit which may have 10% CO.   However, FCCU flue gases  from




    controlled units having complete CO combustion would have very low CO contents




    and would be  satisfactory.





                                   -138-

-------
2.  TELEPHONE CONVERSATION SUMMARIES

     a.  State and Local Air Pollution Control Regulations

     In connection with this study we spoke with representatives of
state and local air pollution control agencies of the states having
a total FCC capacity of 100,000 BPSD.  This represents 71% of total
U.S. capacity, or 102 out of total 143 units.  Questions asked re-
lated to two general areas:

     1.  Confirmation of regulations governing existing sources, and

     2.  Present utilization of CO boilers and trend in the next
         10 years.

Listed below are the answers to these questions from all agencies
contacted.  Information appearing opposite number 3 is additional
information received during conversation.

     Illinois (John Reed)

     1.  Confirmed understanding of SO  and hydrocarbon regulators.
                                      X

     2.  5 of the 9 refineries with FCC's have CO boilers, none
         recently installed.  The trend seems to be more toward
         hot regeneration rather than use of CO boilers.  Although
         HTR is still being studied, Reed thinks that 2 or 3 re-
         fineries presently having variances will move toward this
         process in the next few years.

     3.  Because of Illinois' low sulfur crude requirement, hydro-
         treating has not been popular in the state.  Refineries
         are petitioning for a relaxation of this requirement, however,
         and if they win, hydrotreating might become more popular.

         Many of the refineries have sulfur recovery plants.
     Indiana (Ed Stresino)

     1.  Confirmed understanding of SO  and hydrocarbon regulations,

     2.  2 out of the 4 refineries with FCC's are believed to have
         CO boilers.
                                -139-

-------
2.  TELEPHONE CONVERSATION SUMMARIES Continued
    Kansas (Robert Wallace)

    1.   Confirmed understanding on no source performance standards
        for SO  and hydrocarbons.   There has been hydrocarbon moni-
        toring which led to a decision that no regulation was necessary
        and there is thought of an SO  regulation.
                                     A

    2.   The refinery with an FCC in Kansas County does have a CO boiler.
    Louisiana (Orey Tanner)

    1.   Confirmed understanding of SO  regulation and existence of no
        hydrocarbon regulation.

    2.   Of the 9 FCC facilities in Louisiana, well over half have CO
        boilers.  The Good Hope refinery, which does not have one,
        uses a new DuPont catalyst and claims its CO emissions are low
        enough not to need control.

        Until hearing of this DuPont catalyst recently, Tanner would
        have said the trend was toward increased use of CO boilers.
    New Jersey (Tom Leonard)

    1.   Confirmed understanding of SO  regulation and existence of only
        an ambient air quality standard for hydrocarbons.  Sulfur content
        in fuel regulation is different for commercial and non-commercial
        fuels.  Refineries can be under either classification, but
        generally are considered non-commercial.

    2.   All 4 units in New Jersey have CO boilers.

    3.   Hydrotreating is used by all 4 units in New Jersey.
    Ohio (Howard Johnson)

    1.  Confirmed understanding of SO  regulations and existence of no
        hydrocarbon regulation.

    2.  5 of the 6 units in Ohio have CO boilers.
                                 -140-

-------
2.  TELEPHONE CONVERSATION SUMMARIES Continued

     Oklahoma (John Drake)

     1.  Confirmed understanding that there are no SO  or hydrocarbon
         regulations applicable to FCC's.

     2.  Well over half of the 10 FCC units In the state have them.
     Pennsylvania (Frank Willard)

     1.  Confirmed understanding of SO  regulation and existence of
         no hydrocarbons regulation for FCC's.

     2.  At least 2 of the 4 units In Pennsylvania have CO boilers.
         They are very old and malfunction frequently.

     3.  Both the facilities with CO boilers also have hydrotreating.
     Texas (Frank Spuhler)

     1.  Confirmed understanding of regulations.

     2.  Most of the 30 facilities have CO boilers.  A lesser amount
         use HTR (particularly the newer ones).   Spuhler feels CO
         boiler use is on the increase.


     Washington (Michael Landor)

     1.  Confirmed understanding of SO  regulation and existence of no
         hydrocarbon regulation.
                                 -141-

-------
2.   TELEPHONE CONVERSATION SUMMARIES Continued

     b.  Other Telephone Conversation Summaries

     Davison Chemical (Albert Chatard)

     1.  Data on catalytic cracker feed hydrotreating can be found in
         The Oil and Gas Journal. Annual Refinery Survey.

     2.  Chatard has no information on CO boiler installations.

     3.  An upper limit growth projection for FCC capacity is 2 - 2-1/2%
         per year.  There is considerable uncertainty regarding gasoline
         demand, which is the key to growth.


     Jim Montgomery

     We spoke to Montgomery in the absence of Mr. Blazek, who regularly
     sits on the NPDA Q and A Committee.

     1.  Hydrotreating of FCC feedstock - Montgomery thought about 10%
         of FCC feedstocks are hydrotreated.  He believes the majority
         of this capacity would be in California, due to the high nitrogen
         content of the indigenous crudes.

     2.  CO boilers - Montgomery believes that about 2/3 of all FCC units
         have CO boilers, particularly the larger ones.  This would make
         an aggregate capacity of about 67X.

     3.  FCC capacity growth - Not much information.

     4.  CO promoter catalyst - Montgomery told us that the new catalysts
         promote CO oxidation in the dense bed region where there is an
         adequate heat sink and less chance of local overheating and po-
         tential de-activation of the zeolite catalyst.  Without the pro-
         moter catalyst, the maximum temperature occurs in the dilute
         phase where the metallurgy of the intervals may not be adequate.

     5.  HTR - There are two reasons for applying HTR: 1)  to reduce CO
         emissions to <500 ppm, and 2) to improve catalyst regeneration
         which allows conversion or capacity to increase.  Montgomery
         estimated that there are from 32-34 units using HTR which he
         roughly estimates as 750-900 MBPD of fresh feed capacity.

     6.  Davison catalyst - Approximately 21 units representing about
         300-375 MBPD of capacity, using the Davison catalyst.
                                  -142-

-------
2.   TELEPHONE CONVERSATION SUMMARIES Continued

        and, therefore, must be purged from the system.   A crystallizer
        is used to purge sodium sulfate.   The waste solids are generally
        85% sodium sulfate and 15% sodium sulfite and bisulfite.   Thus,
        the sodium makeup to the Wellman-Lord system should be greater
        than the oxidation plus SO. absorption in the system.

        In the application of the system to utility boilers, the  flue
        gas generally contains very little SO- and has,  on the average,
        4% oxygen.  This combination results in the oxidation of  10%
        of the S0~ absorbed.

        The flue gas from the FCC units contains 1-2% and the SO- varies
        widely.  If the SO- is present in a large quantity and is absorbed
        in the Wellman-Lord system, it will result in large losses of
        sodium.

        The system has very little absorption capacity for NO .
                                -143-

-------
2.   TELEPHONE CONVERSATION SUMMARIES Continued

     Exxon Research and Engineering Company (Terry Day)

     Exxon Is presently updating FCCU capital investment estimates for
     their technical literature.   Previous estimates were $375-450/daily
     barrel of fresh feed, based on second quarter 1973  USGC prices and
     a 35,000 BPD unit.  The estimate includes direct M&L for the
     reactor, regenerator (including blower)  and fractionation.   As a
     rough guess, the regenerator is ~40% of  the total cost.  The total
     erected cost (TEC) would be about 2 times M&L.

     Revamping for HTR in most cases would not cost as much as 50% of
     the TEC of the regenerator.   If a new blower were required, the
     revamp costs could be close to 50%.
     Davy Power Gas (Chris Earl)

     We spoke to Chris Earl about the application of the Wellman-Lord
     system for FCC units»  Most  of the material in connection with the
     operation of the Wellman-Lord system and the economics of the system
     are available in our files.   Therefore,  the conversation was a short
     one and only a few questions were asked.

     1.  The Wellman-Lord technology is available for application.
         Normally, the systems are owned by Davy-Power Gas, and the
         users pay the fee for gas cleanup.   The system has not been
         applied to FCC units; therefore, no  operating information on
         the Wellman-Lord system  on FCC units is available.

     2.  The Wellman-Lord system  is a fully-developed S02 control system
         for commercial application.   There are  about one dozen systems
         either in the planning stage or under construction.   Most of
         these applications are for sulfuric  acid plants or coal- or oil-
         fired utility boilers.  The availability of the system is
         generally high, >90%, and there are  no  major operating problems.

     3.  In the application of the Wellman-Lord  system to the FCC units
         the major question is of oxidation.   The emissions from the FCC
         units may contain from 0-65% of the  SO  , as SO,.   If the SO-
         is present above the condensation point, it will be present as
         a vapor form and will be absorbed in the liquid phase.   However,
         if the SO- is present in the mist form,  it will not  be captured
         in the precleaner of the Wellman-Lord system and will escape
         through the system.   The Wellman-Lord system has a precleaner
         which removes large particles.   The  precleaner has a low pressure
         drop and, therefore, is  not  effective on condensation type of
         nuclei.   The purpose of  the  precleaner  is to ensure  that no
         particles are captured in the S02 absorption section.

         In the Wellman-Lord system,  the absorption of SO, or oxidation
         of sodium sulfite results in the formation of sodium sulfate.
         The sodium sulfate is not regenerated in the Wellman-Lord system

                                 -144-

-------
 APPENDIX D




BIBLIOGRAPHY
    -145-

-------
                              BIBLIOGRAPHY
Air Pollution Engineering Manual, Air Pollution Control District, County
    of Los Angeles, J.A. Danielson (editor), EPA, Office of Air and Water
    Programs, Office of Air Quality Planning and Standards, Research Tri-
    angle Park, N.C., May 1973.

Atmospheric Emissions from Petroleum Refineries, A Guide for Measurement
    and Control, HEW, Public Health Service, Division of Air Pollution,
    Publication No. 763, 1960.

Background Information for Proposed New Source Performance Standards, Vol 1,
    Main Text, EPA, Office of Air & Water Programs, Office of Air Quality
    Planning and Standards, Research Triangle Park, Publication No. APTD-
    1352b, June 1973.

Background Information for Proposed New Source Performance Standards, Vol. 2,
    Appendix:  Summaries of Test Data, EPA, Office of Air and Water Programs,
    Office of Air Quality Planning and Standards, Research Triangle Park,
    Publication No. APTD-1352b, June 1973.

Background Information for New Source Performance Standards, Vol. 3,
    Promulgated Standards, EPA, Office of Air and Water Programs, Office
    of Air Quality Planning and Standards, Research Triangle Park, Publi-
    cation No. AP.TD-1352b, June 1973.

"Complete Combustion of CO in Cracking Process," Chemical Engineering,
    November 24, 1975.

"Converting U.S. Refinery to Sour Crude Costly," The Oil and Gas Journal,
    April 9, 1973.

Ctvrtnicek, T., T. Hughes, C. Moscowitz, and D. Zanders, Monsanto Research
    Corp., "Refinery Catalytic Cracker Regenerator SOX Control Process
    Survey," prepared for Office of Research and Development, U.S. EPA,
    Contract No. 68-02-1320, Task I, Phase I, September 1974.

Cowherd, Chatten, Midwest Research Institute, "Source Testing, EPA Task
    No. 6, Standard Oil of California Company, El Segundo, California,"
    EPA Contract No. 68-02-0228.

Environmental Reporter, State Air Laws, the Bureau of National Affairs,
    Inc., Washington.

Gary, J.H. and H.E. Scheweyer, "Crude Source & Desulfurization," Petroleum
    Refinery, September 1953.

"HDS & FCC Equals More Gasoline," The Oil and Gas Journal, May 17, 1976.
                                  -146-

-------
Ruling, G.P., J.D. McKinney, and T.C. Readal, "Feed Sulfur Distribution in
    FCC Product," The Oil and Gas Journal, May 19, 1975.

Hydrocarbon Processing, 1974 Refining Notebook.

Little, Arthur D., Inc., "The Impact of Lead Additives Regulations  on  the
    Petroleum Refining Industry," EPA Contract No. 68-02-1332,  Task Order
    7, December 1975.

National Emission Data System Point Source Listing (1976 Printout).

The Oil and Gas Journal, Annual Refining Survey, March 29, 1976.

The Oil and Gas Journal, Construction Survey, 1976

Ritter, R.E., J. J. Blazek, and D.N. Wallace, "Hydrotreating FCC  Feed  Could
    be Profitable,"  The Oil and Gas Journal, October 14, 1974.

Schulz, E.J., L. J. Hillenbrand, and R.B. Engdahl, Battelle Columbus
    Laboratories, "Research Report on Source Sampling of Fluid  Catalytic
    Cracking, CO Boiler, and Electrostatic Precipitators at the Atlantic
    Richfield Company, Houston, Texas," EPA Contract No. 68-02-0230,
    Task Order 3, July 6, 1972.

Schulz, E.J., L.J. Hillenbrand, and R.B. Engdahl, Battelle Columbus
    Laboratories, "Research Report on Sampling of Fluid Catalytic
    Cracking Plant (Electrostatic Precipitators and CO Boiler)  of Standard
    Oil of California, Richmond, California," EPA Contract No.  68-02-0230,
    Task Order No. 3, July 6, 1972.

Shea, E.P., Midwest Research Institute, "Source Testing—EPA Task No.  8,
    Standard Oil Company, Richmond, California," EPA Contract No. 68-02-0228.

Shea, E.P., Midwest Research Institute, "Source Testing—EPA Task No.  6,
    Atlantic Richfield Company, Wilmington, California," EPA Contract  No.
    68-02-0228.

"Shell Flue Gas Desulfurization Process Demonstration on Oil-and  Coal-
    Fired Boilers," A.I.Ch.E. Meeting, Tulsa, Oklahoma, March 1974.

"Striking Advances Show Up in Modern FCC Design," The Oil and Gas Journal,
    October 30, 1972.

Wollaston, E.G., "Sulfur Distribution in FCU Products,"The Oil  and  Gas
    Journal, August 2, 1971.
                                  -147-

-------
 APPENDIX E
TRC MODEL IV
     -148-

-------
       Several models have been developed by the EPA for the determination
of priorities over the  past  few years.  The first model provided a compari-
son of source categories based on total atmospheric emissions of all pol-
lutants, availability of control technology, and other factors.  The
second model focused on the  need for individual priorities for each pol-
lutant and attempted to restrict rating criteria to factors selected from
a generalized strategy  for the pollutant.  Impact on emissions was a prime
criterion in all  cases, but  impact was expressed on a relative scale.  The
third model attempted to emphasize impact, but the relative scale concept
was  retained.   Because impact was  expressed in a  relative way,  the model
did  not provide a system amenable  to gradual  refinement  as  available in-
formation was  improved. Other models, developed by EPA, have taken into
account toxicity, exposure,  ambient air  concentrations or population
density.  These models are generally complex  and  not amenable to
modification or refinement in addition to  presenting impact on  a
relative scale.
        Model  IV, which  is  developed below,  is  amenable to data  refinement
and  provides  a quantitative  estimate of  anticipated impact of standards of
performance in preventing  atmospheric emissions.
        The additional  control potential  of new or revised standards of
performance stems from the application of  emission standards that are more
stringent than those presently applied to  construction and modification.
This potential, for a specified time period,  is expressed as
                                   
-------
       To calculate the control  potential  of standards  of performance,
other factors must be considered, such as  the portion of growth  require-
ments that can be satisfied from present unused capacity and the obsole-
scence and replacement rates of existing facilities.   Such a comparison
can be expressed mathematically.
       The following notation is used in the development of the  relation-
ship between projected emissions under baseline year (1975) regulations
and NSPS control.
       TS = total emissions in i   year under baseline year regulations
            (tons/yr)
       Tw = total emissions in i   year under new or revised NSPS which
        n
            have been promulgated in the j   year (tons/yr)
        Ty » total  emissions in  i   year assuming no  control  (tons/yr)
        T. = total  emissions in  baseline year under baseline year regula-
             tions  (tons/yr)
        K  = normal fractional utilization rate of existing capacity,
             assumed constant during time  interval
        A  = baseline year production capacity (production units/yr)
        B  = production capacity from construction and  modification to
             replace obsolete  facilities (production  units/yr)
        C  = production capacity from construction and  modification to  in-
             crease output above baseline  year capacity (production units/yr)
        PB = construction and modification rate to replace obsolete capacity
             (decimal fraction of baseline capacity/yr)
                                  -150-

-------
       PC = construction and modification rate to increase industry capa-
            city (decimal fraction of baseline capacity/yr)
       ES = allowable emissions under existing regulations (mass/unit capa-
            city)
       E» = allowable emissions under standards of performance (mass/unit
            capacity)
       Ey = emissions with no control  (mass/unit capacity)
                                          t"h
       For the purpose of this study the i   year is defined as 1985 and the
3th year, 1975.
       Assuming that capacity lost due to obsolescence is  replaced by con-
struction and modification, as schematically shown in Figure 5-1, then,
       Ts = Es K (A - B) + Es K (B + C)	(1)
and
        TN =  Es K (A -  B) + EN K (B  +  C)	(2)
        Ts*-  TN = K (B  + C) (Es -  EN)	  .   (3)
        Values  of B and C are determined  as  follows:
        (a) If  compound growth is  assumed,
                                                        •
           B = A [ (1  + Pg)1 - 1]	(4)
           C = A [ (1  + Pc)1 - 1]	(5)
        (b) If  simple growth is assumed,
           B = AiPB	(6)
           C = AiPc	(7)
           where
                 i  = elapsed time, years
                                -151-

-------
                 Applicability of NSPS

            to construction and modification
CO

-------
       In addition, the following values may be calculated:



           TA = ESKA .................  (8)



           Ty = EyK (A -B) + EUK (B + C)  ..........  (9)




       Further refinement of the Model may be realized for cases where E~
                                                                        o



for new and existing plants differ.  In this case,



           Tc = KE.  (A - B) + KE_  (B + C)  ......... (10)
            O     J-i             Op




       where:



           Ec  = Ec for existing plants
            b]    b



           Ec  = Ec for new plants
            5     b
       Therefore,
                      K (B + C) (Es  - EN)   ......... (11)
       Section 111 (d) of the Clean Air Act requires the States to regulate


 designated pollutants from existing installations  for industries  to which


 NSPS  have been applied.   To handle this  situation,  a slight modification


 to the model  is necessary.   Ty,  TA and TS are  the  same as for  criteria


 pollutants.  T.. is redefined, however, as
        Where:



                      a^owable  emissions  under State  regulations as re-


                      quired by  Section  lll(d)  of the  Act.


                    = total  emissions  in i   year under Section lll(d) and


                      NSPS regulations.
                                  -153-

-------
       Due to the large number of calculations to be performed and the
repetitive nature of these calculations, the model has been computerized.
A printout of the program can be found in Appendix III.
       For the purposes of this study, K, A, PD and Pr are defined as
                                              D      I
industrial prime variables.  E$  Ey  EN and E^-j ,d» are defined as emis-
sion prime variables.  TA> Ty, T$, TN, TND, B and C are referred to as
intermediate variables.

5.3    INDUSTRIAL FACTORS
5.3.1  Normal Fractional Utilization - K
       The  variable, K, represents that fraction of total existing capa-
city which is brought into service to produce a given output.   By applying
this factor to the capacity based values of A, B and C,  impact on emissions
is determined for actual production.  The numerical value of K may be ar-
bitrarily changed to permit a revaluation of impact on  emissions at any
production level, if so desired.   It is for this reason   that  K exists
within the Model.
       K was generally deduced from information available in the litera-
ture by  simply dividing production  by  capacity.  Occasionally, a direct
reference to the value was made.  Although  the definition appears to be
relatively straightforward,  the  actual  value  can be  interpreted several
ways based upon  the  original  definition of the term  capacity.  Industries
generally specify their production  capacity in two ways—preferred or
physical.   '  Physical capacity  is  defined as the  maximum production that
would  result if  an  industry  pushed  its output to  the ultimate  practical
 level.   Preferred capacity,  on the  other hand,  is  the maximum  quantity of
 production  that  would result considering equipment limitations, normal
 operating schedules, maintenance,  shutdowns and  profits.
                                -154-

-------
       Although the values of K were determined from many data sources
within the literature, several were exceptionally valuable.  The Chemical
                   (5)
Economics Handbook* ' gave production and capacity information for ma0y
industrial categories as well as breakdowns of various processes or methods
within the industry.  A great deal of information necessary to develop K
was obtained from  draft and final reports made available by EPA.  In some
cases, Department  of Commerce Publications, such as the Census of Manufac-
turers* ' and the  Survey of Current Business*  , were used.
       Determination of K for industrial categories that had associated
production and capacity data was relatively straightforward.  A few source
categories, however, did not have an associated production rate and had to
be treated differently.  Field burning of sugar cane, for example, is a
"do or don't" situation.   Therefore, K was set equal to unity for this
type of process.   Emission factors for the dry cleaning industry were
 developed on  a per capita basis with the values of A, B and C being re-
 lated to  population.  Since population is, in fact, "capacity", K was set
equal to  unity.  In general, the majority of manufacturing processes had
a fractional  utilization greater than 0.7.  Most were in the 0.8 to 0.9
range.
5.3.2  Production  Capacity - A
       The variable, A, is defined as industrial production capacity in the
baseline year, 1975.  It is used to derive the values of new (C) or re-
placed (B) capacity in 1985 (Equations 4 through 7) and to define existing
capacity  in 1985 not subject to NSPS (A - B).  Production capacity was
generally determined from production data found in the literature for
                                 -155-

-------
some year other than 1975.  This value was converted to capacity in 1975
by dividing by fractional utilization, K, and scaling this value by PC to
1975.  For those cases where actual capacity was quoted in the literature,
it was not necessary to divide by K.
       The units for capacity were selected to be consistent with those
used by the specific industry and which were compatible with the other
factors.  In most cases, tons of product or tons of feed per year were
chosen.  However, for the phosphate fertilizer industries, tons of  P«05
per year was chosen since production and capacity information is commonly
quoted on this basis.  For combustion sources, A was expressed in BTU's or
horsepower  - years per year.  For incinerators, tons of refuse handled
annually was specified.  Occasionally, the value of capacity was chosen in
terms of a  quantity to which emission or growth factors could be related.
An example  of  this  is degreasing for which tons of metal cleaned, not the
quantity of solvent used, was selected.   Tons of clothes cleaned annually
was chosen as the basis for the dry cleaning analysis.   A growth  rate with-
in the industry based on anticipated population trends  could then be em-
ployed.
       The most recent production or capacity data  available was  used so
that extrapolation to  the baseline year would result  in as  realistic a
value as  possible.   In nearly all  cases,  data sources were  more  recent
than 1967;  much data were from the  1970's.   Several sources were  excep-
tionally  valuable,  notably those mentioned  previously in the discussion
regarding fractional  utilization.  Others included  the  Chemical Profiles
series*  ',  Particulate Pollutant System Study'8', EPA control techniques
documents  and Hydrocarbon Pollutant  Systems  Study'  '.
                                -156-

-------
5.3.3  Increase in Industrial Capacity Over 1975 Capacity -  PC
       The variable, Pp, is defined as the average anticipated growth  rate
in industrial capacity during the period 1975 to 1985.   It is expressed as
a fraction and is applied to A, production capacity, to determine C,
(Equations 5 and 7).  It is this value of C to which NSPS can be applied.
       PC was determined by several methods, the most general being extra-
polation of historical production or capacity data to the year 1985.  A
second relatively common approach  was to relate the anticipated number of
new plants and the average new plant capacity to 1975 capacity levels.  A
third alternative was based on "expert predictions" from sources such  as
Department of Commerce, associated trade associations, cognizant industry
personnel or from studies performed by a number of organizations such  as
the Stanford Research Institute * ' or the Environmental Protection Agency.
 For industries whose function  is directly  related to population, De-
 partment of Commerce data  regarding population  trends '-/as employed to
 determine Pr
            \r»
       A growth  rate based on  "expert predictions" or extrapolation of data
 for the  ten year period, 1975  to 1985, is  subject to the many biases which
 could  occur during  that period.  For example, availability of raw materials,
 sudden changes in demand or consumption patterns  or economic factors such
 as  cost  of money and price controls could  alter historical trends or in-
 validate "expert predictions".  As a result, the  impact of standards would
 be  subsequently  altered.
       As shown  in  Equations 5 and 7, Pp may be expressed as either a
 compound or simple  growth  rate.  If the historical growth pattern was  in-
 deed compound  in nature,  the  value of Pp  was  calculated by the following
 equation:
                                 -157-

-------
       P  -  x"y i  Capacity  in year  "x"    , n
       PC "         Capacity  in year  "y"     l>u
       where x>y
       If the historical  growth  pattern was simple in nature, the value of
Pr  was calculated by the following equation:
 w
       D   - Capacity in  year "x"  - Capacity in year "y"   ....   (14)
       KC             (x-y)  Capacity  in 1975
       where x>y
       For the case of simple growth, it  is necessary to relate the growth
to the baseline year, 1975,  as shown  above.  For compound  growth, the  rate
can be applied to any year.
       For the majority of cases,  P-  was  approximated by a compound rate
and based on the most recent data  available to preclude major inaccuracies
in the determination of emission impact,
       There were several cases  where the anticipated growth rate exceeded
105 annually.  Industries characterized by both a rapid growth rate and
significant emission rates are prime  candidates for NSPS since almost
conplete control of the industry can  be recognized in a relatively short
tine period.  Since this  rapid growth rate will eventually level off,  the
use of the Model should not  be extrapolated  too far beyond the baseline
year or an unrealistic value of  emission  impact could result.  Future
growth rates in these industries should,  therefore, be carefully
mon'tored.
                                  -158-

-------
      For several industries such as lead pigment manufacture  or ROP
triple superphosphate production, a continuing downward trend  in capa-
city was noted.  This characteristic was associated with industries
being phased out due to replacement by more efficient processes, or the
demand for whose product was declining because of the availability of a
better or cheaper product.  For the purpose of our study, we assumed
that these industries did not replace obsolete facilities due  to the
lack of economic incentive.  Accordingly, there would be no  new or
modified capacity generated between 1975 and 1985 that could be controlled
by NSPS.  However, under Section Hid of the Clean Air Act,  the States are
required to regulate designated pollutants from existing installations
for industries to which NSPS have been applied.  It was necessary, there-
fore, to determine TS,  TND and (TS-TN[))  for those sources with decreasing
production capable of emitting designated pollutants.  Values  for the
year 1985 are included in Section 6.0, Presentation of Results.  In
addition, values for each year between 1975 and 1985 are presented in
Section 7.0, Analysis.  This has been done since the emission  impact  is
greatest  in  1976 and diminishes  throughout the ten year period due to the
decreasing capacity.
        For several  other  industries,  a  zero growth  rate was observed.  We
assumed,  however, that obsolete  facilities were  replaced, thereby permit-
                        •
ting an emission impact calculation to  be performed on  the value  B,  the
obsolete  production capacity  replaced between  1975  and  1985.
                               -159-

-------
5.3.4  Replacement Rate of Obsolete Production  Capacity  -  PB
       The variable, Pg, is defined as the average  rate  at which  obsolete
production capacity is replaced during the period 1975 to  1985.   It  is
expressed as a fraction and is applied to A to  determine B, (Equations  4
and 6).  It is this value of B to which NSPS can  be applied.  Also,  the
quantity, (A-B), defines the existing production  capacity  in  1985 to which
only State regulations are applicable.
       PB was determined by one of three methods.  One approach was  to  re-
late the number of known or estimated plant closings and the  average exist-
ing plant capacity to 1975 capacity levels.  A  second method  was  based  on
known equipment lifetime.  For example, if a major  piece of production
equipment had an actual estimated lifetime of 50  years,  it would  depreciate
at a rate of 2% per year on a simple basis.  The  third,  and most  common,
method was to use depreciation guidelines published by the Internal  Revenue
Service    .  The allowance permitted by the IRS  is an economic factor  used
for tax collection purposes and generally depreciates equipment and  facili-
ties over a shorter term than their actual useful life.  We assumed  for the
purpose of this study, therefore, that typical  equipment and  facilities
within each industrial category evaluated had a useful life equal to twice
that allowed by the IRS.  As a general rule, P  was based  on  very limited
                                             B
data and, as a result, a great deal of judgment was necessary.  For  this
reason, Pg  was selected on the basis of straight line depreciation  (simple)
to avoid compounding potential errors.
                                  -160-

-------
5.4-    EMISSION FACTORS

5.4.1  Uncontrolled Emission Factor - Ey
       The variable, Ey, is the emission factor representing a condition of
no control.  It is used to calculate Ty, the uncontrolled emissions in 1985,
the value to which TS and TN may be compared to determine the nationwide
impact on emissions of regulations in general.  Ey is also employed to de-
velop EN, the NSPS controlled emission factor.  When the efficiency of a
control device is stated, application of this efficiency to Ey  results in
the calculation of E»   Thirdly, Ey replaces the value of E<., the emission
factor representing control to the extent required by State regulations,
when no regulations for an industry exist in a given state.
       Ey, in most cases, represents a totally uncontrolled emission factor.
On occasion, however, it represents the controlled emission factor at the
exit of a control device if such a device is integral with the process.  An
example of this would be carbon black manufacture by the furnace process
where the product is actually collected by a series of control devices.  If
these devices were not functional, the process could not operate.
       Determination of Ey was relatively straightforward and references in
the literature were abundant.  Compilation of Air Pollutant Emission Factors,'
AP-42   ' and Air Pollutant Emission Factors*  ' were major reference
sources for this value.  Uncontrolled emission factors for particulates were
                                                                        /0\
determined for many sources from the Particulate Pollutant Systems Studyv '.
Occasionally, E.. for a specific process or operation was synthesized from
several independent values of Ey.  Examples were fossil fuel fired boilers,
gas turbine engines and internal combustion engines where the value of E..
was determined by weighting the emission factors for each fuel type by the
                                    -161-

-------
fraction of the total heating value supplied by each  fuel.   For certain
industries, Ey was synthesized by weighting the emission factors from
different portions of an operation by the fraction of total  capacity as-
sociated with that operation.  It is for this reason  that the value of Ey
developed for this study should not be used in any other context or erro-
neous conclusions could result.
       Units for Ey were chosen to be consistent with the units selected
for A, production capacity.  For example, if A were in terms of tons of
product per year, Ey would be specified in terms of pounds  of emissions
per ton of product.  For those cases where literature quotations for Ey
were on a different basis than A, it was necessary to make  the proper con-
version.  Generally speaking, emission factors for fugitive emissions were
not included within the study even though for certain industries fugitive
emissions may be greater than emissions from point sources.    This is an
area where further study is necessary to quantitatively assess the impor-
tance of this category of sources and to develop methods for emission
control.
5.4.2  Controlled Emission Factor - EN
       The variable, E^, is the emission factor representing the condition
of best control applied to new sources.  It is used to determine TN, the
emissions that would exist in 1985 if NSPS were applied.  When TN is sub-
tracted from TS, the quantitative value of emission impact is determined.
The units of EN were chosen to be consistent with those selected for E...
       A literature search was conducted to find the  best level of control
that could be applied to new or modified construction.  The information
                                 -162-

-------
came from a wide variety of references.  In addition to those mentioned
for Ey, which occasionally gave controlled emission factors, the IGCI
surveys*  *   ' and feature articles and process summaries from various
trade magazines such as Chemical Engineering were exceptionally valuable.
The determination of EN was accomplished by one of three methods.  The
first, and most common, was directly from information regarding a well-
controlled plant.  The second method was by applying a stated control
hardware efficiency to the value of Ey.  When no reference to control tech-
niques was made, a transfer of technology from similar processes was assu-
med where it was deemed applicable.
       There were several instances where technology to control a specific
pollutant within an Industry had not been demonstrated and for which a
transfer of technology was judged not feasible in our opinion due to
technical or economic reasons or for which no control efforts were ever
made due to a low associated point source emission rate.  For those cases,
EN was developed by assuming the anticipated level of control that would
result if present research and development efforts are successful.  For
those cases where no specific research and development efforts are present-
ly underway, we set EN = 0.0 to determine the maximum hypothetical impact
on emissions if the pollutant were to be completely controlled.  The pur-
pose of this application was to develop a separate listing of industrial
categories, ranked in order of hypothetical emission impact, from which
priorities for control technology research and development efforts can be
developed.  The results are weighted towards categories for which there are
no present control efforts (E^ = 0.0).  They should not be compared to the
values determined for the majority of cases where control technology has
                                  -163-

-------
been demonstrated or where a transfer of technology was  judged feasible.
This listing is therefore presented separately in  Tables 6-1  through  6-13.
       The values of EN determined for this study  were based  on present
levels of control technology as determined from the literature.  It is
possible, however, that nationwide plant surveys for each industry could
locate more efficient techniques for unique installations which have  not
been presented in the literature.  Such an effort  was beyond  the scope  of
this project.  It is also recognized that advancements in the state-of-the-
art of control technology will occur and the value of E.., will consequently
change as time goes On.
5.4.3  Controlled Emission Factor For Designated Pollutants - E^j^)

       The variable, ^w^)* is the emission factor which represents best
                             *
control applied to designated  pollutants from existing  plants.  It is  used
to determine T.-D, the emissions of a designated pollutant that would  exist
in 1985 for existing plants under State control and new  plants under  NSPS.
Section lll(d) of the Clean Air Act^2'3^ requires  the States  to draft,
maintain and enforce regulations for the control of designated pollutants
from existing sources for which NSPS have been set for new sources within
that industry.  As a result, the Model was modified to reflect this situa-
tion (see Equation 12).  When TND is subtracted from T<,, the  quantitative
*For the purpose of this study the following  pollutants are defined  as
designated:   fluorides,  trace  metals, acid mist, lead, ammonia, sulfides,
chlorine and odors.

                               -164-

-------
value of emission impact for designated pollutants is determined.  Units
are identical to E,,.
       Determination of E-    was accomplished in a manner similar to EN
Control techniques or levels, however, differed in that retrofit technology
was necessary for existing plants.  In most cases, we believe that avail- '
able control technology for new installations could also be retrofit to
existing installations .  Of course, there would be specific instances at
certain individual plants where this might not be possible due to existing
structures and prohibitive costs.
       Equation  (12) was also used to calculate controlled emissions of
"hazardous" pollutants as defined in Section 112 of the Clean Air Act.   How-
ever, E-nKd) was replaced by EN» since both new and existing sources would
be controlled.
5.4.4  Estimated Allowable Emission Under 1975 Regulations - E_
       The variable, ES, is the emission factor which represents the 1975
level of control required under State, local, regional or Federal regula-
tions.  It is used to determine T-, emissions in 1985 under baseline year
regulations.  When T.. is subtracted from TS, the quantitative impact of
NSPS on emissions is calculated.
       To determine the applicable regulations, tabulations were made on a
State by State, pollutant by pollutant, and industry by industry (where ap-
plicable) basis.  This was done by updating and augmenting all summary
tables published in Analysis of Final State Implementation Plans'15^ (to
August 1974) by reviewing all the State regulations as published in the
Environment Reporter*   .  Federal regulations for new sources promulgated
under Section 111 of the Clean Air Act were also incorporated.  Ready
access to all regulations was thereby provided in tabular form.  No effort
was mads to account for anticipated state regulations beyond 1975.
                                -165-

-------
       The value of E~ was usually a weighted average of all existing regu-
 lations and/or Ey if no regulations existed.  Weighting was generally deter-
 mined on the basis of production capacity distribution as outlined below.
                  N
                          A
                           i   .....   .........
                i  = 1
       where:
       Ec   =  Regulation in 4th State.
        bi
       A.   =  Decimal fraction of total  capacity located in 1th 'State.
       i    =  Individual State
       N    =  Total number of States over which capacity is distributed
Since the units of Ec and E_  had to be  the  same as those chosen for E,.
                    o      o-                                         U
and EN> it was necessary to convert the  majority of regulations into the
appropriate units.  For example, a particulate emission regulation in
pounds per hour was converted to pounds  per ton by dividing by the typical
plant size on  a tons per hour basis.
       Determination of E- for parti cul ate emissions from general processes
required a series of calculations.  For most States, a process weight curve
constitutes the regulation.  To determine the allowable emissions in pounds
per hour for each State, we first determined the process weight rate in tons
per hour for a typical plant in that State.   This was done by dividing the
production capacity in the State by the  number of associated plants.  This
was in turn converted from an annual rate to an hourly rate by applying the
number of annual operating hours.  Since this value was generally in terms
of output capacity, we then developed a feed to product ratio which con-
                                -166-

-------
verted output to input capacity—the value required by the process weight
rate curve.  The allowable emission in pounds per hour from the curve was
divided by the output capacity to obtain Ec  in proper units.   When these
                                          bi
calculations had been made for each State of concern, the average value of
ES was determined as described by Equation (15).  This is a typical example
of ES determination.  There were many variations and special case situations
too numerous to mention here.  Details of the various calculations for each
industrial category can be found in the appropriate Appendices.
       There were many cases where sources of particulate emissions were
distributed throughout all States and specific data regarding  geographical
distribution was lacking.  For these cases, we developed a generalized
process weight rate curve by linearly averaging the process weight curves
of all the States at each production level.  The resulting curve is pre-
sented in Figure 5-2.  It is slightly less stringent than the  curve origi-
nally issued as the EPA guideline and slightly more stringent  than a curve
generated by weighting the process weight curves of the twenty-five most
populated States on the basis of fractional population distribution.  Sim-
ilar generalizations were developed for the particulate, NO  and SO  regu-
                                                           A       /\
lations for fuel burning sources.
       For hydrocarbon emission sources, the value of E$ is related to the
reactivity of the pollutant since different regulations apply  to reactive
and non-reactive hydrocarbons.  For the case of reactive emissions, we
determined the typical plant size within each state of concern, calculated
the hourly emission rate and applied the "percent control" regulation, if
it existed, to determine whether the hourly or "percent control" emission
regulation was applicable.  After converting the result for each State to
                               -167-

-------
102 I-
              ;  xlO1   ?     ?   {  ?  f ? f ?! xlO2   ?     ?   f  ?  !  ? ?  ? |x103   ?     ?   1  ?  I  f ? I
                                                9getMii!iiinmtjuui|!i^
   e:
   a:
   ^
   CO
   CO
  CO
  o

-
  LU

  co


  c
  _J
  _J
  <
10'
                                  AVERAGE OF  STATE REGULATIONS

                                      APPLICABLE  TO GENERAL

                                  PROCESS SOURCES  -  PARTICULATES



                                                FIGURE 5-2
                                                                                                            I   «  ?  ! ? • «!xlO
                                                                                                                                                                        x!0c
                                                                                                     UlinilHIMWHIWIIIIIIHItlllHIlUlllUtiinBBIIIUIIIIUIiniUIUMIUIilUUUUtUUIIUtlUllt
                                                                                             ____,,,lliJISsJil|!Hiillll!ill!li^itilli!llllil;:;«i:,gSfeBSiS533!5M!fi!!JliiHaii!S!liait!lHlil(Ninf
                                                                                             e£3f S^^^^^

                                                                                                        	"		"	"""
        .f: rrmrr
        I
i
                    illlllililil'iliilillinlll'ililHiliiHiiig
                                                                                                ^

                               lpHinUlllllNllllllil»!Hlll|KaillllllHlllll!!!IIIIIMIIIIIIIItflllll!lllij|[lllllllII!IIUnillllllUlliUlllllin!l!llilll!lltt
              •iiiiiiiiiiiuiiuiiuiiiimiiiiiiiH
              •lllllimitUinUWUilliilill!III
-------
pounds per tora, a weighting process, as shown in Equation (15) was perform-
ed.  As was the case for general process parti oil ate emission ES determina-
tion, there were many variations and special case considerations too numer-
ous to mention here.  Details of the various calculations for each indus-
trial category can be found in the appropriate Appendices.
       For several cases, State regulations for new sources differed from
existing sources.  Consequently, two values of ES were determined and ap-
plied to the Model as shown in Equation (10).
       Regulations pertaining to visible emissions were not included in the
evaluation of E- due to the  impracticality of converting stated opacity  levels
to a weight rate of emissions for use in the Model.
                                    -169-

-------
                     APPENDIX F
RELATIONSHIP OF PAD DISTRICTS TO AIR QUALITY CONTROL REGIONS
                        -170-

-------
     In this study we have looked at the U.S. FCC capacity in terms of
PAD Districts of which there are 5 (see Figure F-l).   Another way in which
the FCC population can be categorized is by Air Quality Control Regions,
of which there are 10 (see Figure F-2).  The relationships between these
two categorizations are shown by state in Table F-l.
                                 -171-

-------
 PETROLEUM ADMINISTRATION FOR DEFENSE (PAD) DISTRICTS
    (
    r
(Incl Alaska
arid Hawaii)
 \   (  •**
                        FIGURE F-l

-------
            UNITED STATES
ENVIRONMENTAL PROTECTION AGENCY
             Regional Offices
                                         r\
                 FIGURE F-2

-------
                               TABLE F-l
                      U. S. PETROLEUM REFINERIES

                         WITH FCC REGENERATORS
State
  No. of
Ref. w/FCC
Total Capacity
    (b/sd)
(January 1, 1976)
  ACQR       PAD
Arkansas
California
Colorado
Delaware
Hawaii
Illinois
Indiana
Kansas
Kentucky
Louisiana
Michigan
Minnesota
Mississippi
Missouri
Montana
Nebraska
New Jersey
New Mexico
New York
North Dakota
Ohio
Oklahoma
Pennsylvania
Tennessee
Texas
Utah
Virginia
Washington
Wisconsin
Wyoming
     1
    13
     2
     1
     1
     9
     4
    10
     1
     9
     3
     3
     2
     1
     4
     1
     4
     2
     2
     1
     6
    10
     4
     1
    30
     6
     1
     3
     1
   	7
   143
18,000
621,211
23,900
77,000
23,000
523,277
203,800
206,450
55,000
677,728
46,900
74,500
76,850
53,000
72,500
2,900
269,444
17,560
47,000
34,000
248,500
231,675
224,300
13,500
1,527,571
70,160
32,000
118,600
10,700
74,078
VI
IX
VIII
I
IX
V
V
VII
IV
VI
V
V
IV
VII
VIII
VII
II
VI
II
VIII
V
VI
III
IV
VI
X
III
X
V
VIII
3
5
4
1
5
2
2
2
2
3
2
2
3
2
4
2
1
3
1
2
2
2
1
2
3
4
1
5
2
4
                                       5,675,104
Source:  Oil and Gas Journal, March  29,  1976.
                                  -174-

-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1.
4.
REPORT NO. 2.
EPA-450/3-77-046
TITLE AND SUBTITLE
Screening Study to Determine Need for SO
Hydrocarbon NSPS for FCC Regenerators

and
7. AUTHOR(S)
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts 02140
12
15
16
17
a.
. SPONSORING AGENCY NAME AND ADDRESS
U. S. Environmental Protection Agency
Research Triangle Park, North Carolina 2
. SUPPLEMENTARY NOTES
7711
3. RECIPIENT'S ACCESSION^NO.
5. REPORT DATE
August 1976
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1 1 . CONTRACT/GRANT NO.
68-02-1332, Task 22
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
200/04

. ABSTRACT
The report quantifies the emissions from a typical Fluid Catalytic Cracking (FCC)
unit and identifies and presents all available data that would define the emission
levels that can be achieved with the most effective demonstrated control systems
for FCC units. In addition, estimated emission reductions that would result
through promulgation of new source performance standards for FCC units have been
prepared.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution Fluid Catalytic
Emissions Cracking
Control Equipment Regenerator
Sulfur Dioxide
Carbon Monoxide
Hydrocarbons
Petroleum Refining
18
DISTRIBUTION STATEMENT
Unlimited
b. IDENTIFIERS/OPEN ENDED TERMS C. COS ATI Field/Group
Air Pollution Content
Stationary Sources
Hydrocarbon Emission
Control
Sulfur Dioxide Control
Carbon Monoxide Control
19. SECURITY CLASS (This Report) 21. NO. OF PAGES
Unclassified 187
20. SECURITY CLASS (This page) 22. PRICE
Unclassified
EPA Form 2220-1 (9-73)
                                                     -175-

-------